owencqueen commited on
Commit
34fd117
·
verified ·
1 Parent(s): 3c3b1b8

Upload 228 files

Browse files
This view is limited to 50 files because it contains too many changes.   See raw diff
Files changed (50) hide show
  1. VCI/parsing_csr_criteria/cspec_version_guide.csv +253 -0
  2. VCI/parsing_csr_criteria/cspec_version_guide_processed.csv +0 -0
  3. VCI/parsing_csr_criteria/version_csv_individual/ClinGenACADVLExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv +255 -0
  4. VCI/parsing_csr_criteria/version_csv_individual/ClinGenBrainMalformationsExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1.1.0_version=1.1.0.csv +104 -0
  5. VCI/parsing_csr_criteria/version_csv_individual/ClinGenBrainMalformationsExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv +81 -0
  6. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv +108 -0
  7. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion3.1_version=3.1.0.csv +101 -0
  8. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion3_version=3.0.0.csv +101 -0
  9. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforCDH1Version1.0.0_version=1.0.0.csv +101 -0
  10. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv +72 -0
  11. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACTC1Version1.0.0_version=1.0.0.csv +871 -0
  12. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYBPC3Version1.0.0_version=1.0.0.csv +952 -0
  13. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYH7Version2.0.0_version=2.0.0.csv +900 -0
  14. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYL2Version1.0.0_version=1.0.0.csv +882 -0
  15. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYL3Version1.0.0_version=1.0.0.csv +882 -0
  16. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTNNI3Version1.0.0_version=1.0.0.csv +897 -0
  17. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTNNT2Version1.0.0_version=1.0.0.csv +897 -0
  18. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTPM1Version1.0.0_version=1.0.0.csv +898 -0
  19. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGAMTVersion1.1.0_version=1.1.0.csv +324 -0
  20. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGAMTVersion1_version=1.0.0.csv +610 -0
  21. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGAMTVersion2.0.0_version=2.0.0.csv +406 -0
  22. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGATMVersion1.1.0_version=1.1.0.csv +337 -0
  23. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGATMVersion1_version=1.0.0.csv +568 -0
  24. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGATMVersion2.0.0_version=2.0.0.csv +272 -0
  25. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSLC6A8Version1.1.0_version=1.1.0.csv +209 -0
  26. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSLC6A8Version1.2.0_version=1.2.0.csv +208 -0
  27. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSLC6A8Version1_version=1.0.0.csv +190 -0
  28. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCoagulationFactorDeficiencyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforF8Version1.0.0_version=1.0.0.csv +122 -0
  29. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCoagulationFactorDeficiencyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforF9Version1.0.0_version=1.0.0.csv +111 -0
  30. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACTA1Version1.0.0_version=1.0.0.csv +180 -0
  31. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACTA1Version2.0.0_version=2.0.0.csv +163 -0
  32. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDNM2Version1.0.0_version=1.0.0.csv +161 -0
  33. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMTM1Version1.0.0_version=1.0.0.csv +182 -0
  34. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforNEBVersion1.0.0_version=1.0.0.csv +218 -0
  35. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRYR1Version1.0.0_version=1.0.0.csv +187 -0
  36. VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRYR1Version2.0.0_version=2.0.0.csv +185 -0
  37. VCI/parsing_csr_criteria/version_csv_individual/ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1.1.0_version=1.1.0.csv +205 -0
  38. VCI/parsing_csr_criteria/version_csv_individual/ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1.2.0_version=1.2.0.csv +139 -0
  39. VCI/parsing_csr_criteria/version_csv_individual/ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1.3.0_version=1.3.0.csv +139 -0
  40. VCI/parsing_csr_criteria/version_csv_individual/ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1_version=1.0.0.csv +212 -0
  41. VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA1Version1.0.0_version=1.0.0.csv +373 -0
  42. VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA1Version1.1.0_version=1.1.0.csv +373 -0
  43. VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA1Version1.2.0_version=1.2.0.csv +364 -0
  44. VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA2Version1.0.0_version=1.0.0.csv +373 -0
  45. VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA2Version1.1.0_version=1.1.0.csv +373 -0
  46. VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA2Version1.2.0_version=1.2.0.csv +364 -0
  47. VCI/parsing_csr_criteria/version_csv_individual/ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1AVersion1.0.0_version=1.0.0.csv +521 -0
  48. VCI/parsing_csr_criteria/version_csv_individual/ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1AVersion2.0.0_version=2.0.0.csv +461 -0
  49. VCI/parsing_csr_criteria/version_csv_individual/ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1BVersion1.0.0_version=1.0.0.csv +506 -0
  50. VCI/parsing_csr_criteria/version_csv_individual/ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1BVersion2.0.0_version=2.0.0.csv +458 -0
VCI/parsing_csr_criteria/cspec_version_guide.csv ADDED
@@ -0,0 +1,253 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ ,Title,Genes,Version,Released Date,Link,path
2
+ 0,ClinGen Cardiomyopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MYH7 Version 2.0.0,MYH7,2.0.0,4/22/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN002?version=2.0.0,ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYH7Version2.0.0_version=2.0.0.csv
3
+ 1,ClinGen Cardiomyopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,MYH7,1.0.0,6/25/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN002?version=1.0.0,ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
4
+ 0,ClinGen PTEN Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PTEN Version 3.1.0,PTEN,3.1.0,3/14/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN003?version=3.1.0,ClinGenPTENExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPTENVersion3.1.0_version=3.1.0.csv
5
+ 1,ClinGen PTEN Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PTEN Version 3.0.0,PTEN,3.0.0,3/27/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN003?version=3.0.0,ClinGenPTENExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPTENVersion3.0.0_version=3.0.0.csv
6
+ 2,ClinGen PTEN Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2,PTEN,2.0.0,9/10/2019,https://cspec.genome.network/cspec/ui/svi/doc/GN003?version=2.0.0,ClinGenPTENExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv
7
+ 3,ClinGen PTEN Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PTEN Version 1.0.0,PTEN,1.0.0,8/17/2018,https://cspec.genome.network/cspec/ui/svi/doc/GN003?version=1.0.0,ClinGenPTENExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPTENVersion1.0.0_version=1.0.0.csv
8
+ 0,"ClinGen Hearing Loss Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for CDH23, COCH, GJB2, KCNQ4, MYO6, MYO7A, SLC26A4, TECTA and USH2A Version 2","CDH23, COCH, GJB2, KCNQ4, MYO6, MYO7A, SLC26A4, TECTA, USH2A",2.0.0,3/30/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN005?version=2.0.0,"ClinGenHearingLossExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforCDH23,COCH,GJB2,KCNQ4,MYO6,MYO7A,SLC26A4,TECTAandUSH2AVersion2_version=2.0.0.csv"
9
+ 1,ClinGen Hearing Loss Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"TECTA, KCNQ4, SLC26A4, MYO7A, USH2A, MYO6, GJB2, COCH, CDH23",1.0.0,8/15/2018,https://cspec.genome.network/cspec/ui/svi/doc/GN005?version=1.0.0,ClinGenHearingLossExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
10
+ 0,ClinGen Phenylketonuria Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PAH Version 2.0.0,PAH,2.0.0,7/16/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN006?version=2.0.0,ClinGenPhenylketonuriaExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPAHVersion2.0.0_version=2.0.0.csv
11
+ 1,ClinGen PAH Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,PAH,1.0.0,3/2/2018,https://cspec.genome.network/cspec/ui/svi/doc/GN006?version=1.0.0,ClinGenPAHExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
12
+ 0,ClinGen CDH1 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 3.1,CDH1,3.1.0,3/29/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN007?version=3.1.0,ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion3.1_version=3.1.0.csv
13
+ 1,ClinGen CDH1 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 3,CDH1,3.0.0,9/21/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN007?version=3.0.0,ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion3_version=3.0.0.csv
14
+ 2,ClinGen CDH1 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2,CDH1,2.0.0,9/6/2019,https://cspec.genome.network/cspec/ui/svi/doc/GN007?version=2.0.0,ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv
15
+ 3,ClinGen CDH1 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for CDH1 Version 1.0.0,CDH1,1.0.0,9/19/2018,https://cspec.genome.network/cspec/ui/svi/doc/GN007?version=1.0.0,ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforCDH1Version1.0.0_version=1.0.0.csv
16
+ 0,ClinGen Myeloid Malignancy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2,RUNX1,2.0.0,9/15/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN008?version=2.0.0,ClinGenMyeloidMalignancyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv
17
+ 1,ClinGen Myeloid Malignancy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,RUNX1,1.0.0,7/10/2019,https://cspec.genome.network/cspec/ui/svi/doc/GN008?version=1.0.0,ClinGenMyeloidMalignancyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
18
+ 0,ClinGen TP53 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TP53 Version 2.3.0,TP53,2.3.0,2/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN009?version=2.3.0,ClinGenTP53ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTP53Version2.3.0_version=2.3.0.csv
19
+ 1,ClinGen TP53 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TP53 Version 2.2.0,TP53,2.2.0,9/30/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN009?version=2.2.0,ClinGenTP53ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTP53Version2.2.0_version=2.2.0.csv
20
+ 2,ClinGen TP53 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TP53 Version 2.1.0,TP53,2.1.0,9/13/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN009?version=2.1.0,ClinGenTP53ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTP53Version2.1.0_version=2.1.0.csv
21
+ 3,ClinGen TP53 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TP53 Version 2.0.0,TP53,2.0.0,7/30/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN009?version=2.0.0,ClinGenTP53ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTP53Version2.0.0_version=2.0.0.csv
22
+ 4,ClinGen TP53 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TP53 Version 1.4.0,TP53,1.4.0,7/5/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN009?version=1.4.0,ClinGenTP53ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTP53Version1.4.0_version=1.4.0.csv
23
+ 5,ClinGen TP53 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TP53 Version 1.3.0,TP53,1.3.0,3/8/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN009?version=1.3.0,ClinGenTP53ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTP53Version1.3.0_version=1.3.0.csv
24
+ 6,ClinGen TP53 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1.2,TP53,1.2.0,8/6/2019,https://cspec.genome.network/cspec/ui/svi/doc/GN009?version=1.2.0,ClinGenTP53ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1.2_version=1.2.0.csv
25
+ 7,ClinGen TP53 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TP53 Version 1.0.0,TP53,1.0.0,8/6/2019,https://cspec.genome.network/cspec/ui/svi/doc/GN009?version=1.0.0,ClinGenTP53ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTP53Version1.0.0_version=1.0.0.csv
26
+ 0,ClinGen Lysosomal Storage Disorders Variant Curation Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2,GAA,2.0.0,6/2/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN010?version=2.0.0,ClinGenLysosomalStorageDisordersVariantCurationExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv
27
+ 0,ClinGen Platelet Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2.1,"ITGA2B, ITGB3",2.1.0,11/1/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN011?version=2.1.0,ClinGenPlateletDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2.1_version=2.1.0.csv
28
+ 1,ClinGen Platelet Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2.0,"ITGA2B, ITGB3",2.0.0,9/4/2020,https://cspec.genome.network/cspec/ui/svi/doc/GN011?version=2.0.0,ClinGenPlateletDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2.0_version=2.0.0.csv
29
+ 2,ClinGen Platelet Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ITGA2B Version 1.0.0,"ITGA2B, ITGB3",1.0.0,6/12/2020,https://cspec.genome.network/cspec/ui/svi/doc/GN011?version=1.0.0,ClinGenPlateletDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforITGA2BVersion1.0.0_version=1.0.0.csv
30
+ 0,ClinGen Malignant Hyperthermia Susceptibility Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RYR1 Version 2,RYR1,2.0.0,3/1/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN012?version=2.0.0,ClinGenMalignantHyperthermiaSusceptibilityExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRYR1Version2_version=2.0.0.csv
31
+ 1,ClinGen Malignant Hyperthermia Susceptibility Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,RYR1,1.0.0,11/9/2020,https://cspec.genome.network/cspec/ui/svi/doc/GN012?version=1.0.0,ClinGenMalignantHyperthermiaSusceptibilityExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
32
+ 0,ClinGen Familial Hypercholesterolemia Expert Panel Specifications to the ACMG/AMP Variant Classification Guidelines Version 1.2,LDLR,1.2.0,11/9/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN013?version=1.2.0,ClinGenFamilialHypercholesterolemiaExpertPanelSpecificationstotheACMGAMPVariantClassificationGuidelinesVersion1.2_version=1.2.0.csv
33
+ 0,ClinGen Mitochondrial Disease Nuclear and Mitochondrial Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1_ntDNA,"SLC19A3, PDHA1, POLG, ETHE1",1.0.0,4/30/2020,https://cspec.genome.network/cspec/ui/svi/doc/GN014?version=1.0.0,ClinGenMitochondrialDiseaseNuclearandMitochondrialExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_ntDNA_version=1.0.0.csv
34
+ 0,ClinGen Mitochondrial Disease Nuclear and Mitochondrial Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1_mtDNA,,1.0.0,4/30/2020,https://cspec.genome.network/cspec/ui/svi/doc/GN015?version=1.0.0,ClinGenMitochondrialDiseaseNuclearandMitochondrialExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_mtDNA_version=1.0.0.csv
35
+ 0,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for HNF1A Version 2.1.0,HNF1A,2.1.0,8/11/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN017?version=2.1.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforHNF1AVersion2.1.0_version=2.1.0.csv
36
+ 1,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for HNF1A Version 2.0.0,HNF1A,2.0.0,1/11/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN017?version=2.0.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforHNF1AVersion2.0.0_version=2.0.0.csv
37
+ 2,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1.2,HNF1A,1.2.0,6/5/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN017?version=1.2.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1.2_version=1.2.0.csv
38
+ 3,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1.1,HNF1A,1.1.0,6/5/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN017?version=1.1.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1.1_version=1.1.0.csv
39
+ 4,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for HNF1A Version 1.0.0,HNF1A,1.0.0,6/4/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN017?version=1.0.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforHNF1AVersion1.0.0_version=1.0.0.csv
40
+ 0,ClinGen Brain Malformations Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1.1.0,"AKT3, MTOR, PIK3CA, PIK3R2",1.1.0,8/19/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN018?version=1.1.0,ClinGenBrainMalformationsExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1.1.0_version=1.1.0.csv
41
+ 1,ClinGen Brain Malformations Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"AKT3, MTOR, PIK3CA, PIK3R2",1.0.0,5/15/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN018?version=1.0.0,ClinGenBrainMalformationsExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
42
+ 0,ClinGen Glaucoma Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MYOC Version 2.0.0,MYOC,2.0.0,12/12/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN019?version=2.0.0,ClinGenGlaucomaExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYOCVersion2.0.0_version=2.0.0.csv
43
+ 1,ClinGen Glaucoma Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1.1,MYOC,1.1.0,11/19/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN019?version=1.1.0,ClinGenGlaucomaExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1.1_version=1.1.0.csv
44
+ 2,ClinGen Glaucoma Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,MYOC,1.0.0,10/12/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN019?version=1.0.0,ClinGenGlaucomaExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
45
+ 0,"ClinGen Hereditary Breast, Ovarian and Pancreatic Cancer Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ATM Version 1.3.0",ATM,1.3.0,3/27/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN020?version=1.3.0,"ClinGenHereditaryBreast,OvarianandPancreaticCancerExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforATMVersion1.3.0_version=1.3.0.csv"
46
+ 1,"ClinGen Hereditary Breast, Ovarian and Pancreatic Cancer Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ATM Version 1.2.0",ATM,1.2.0,11/28/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN020?version=1.2.0,"ClinGenHereditaryBreast,OvarianandPancreaticCancerExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforATMVersion1.2.0_version=1.2.0.csv"
47
+ 2,"ClinGen Hereditary Breast, Ovarian and Pancreatic Cancer Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ATM Version 1.1",ATM,1.1.0,2/25/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN020?version=1.1.0,"ClinGenHereditaryBreast,OvarianandPancreaticCancerExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforATMVersion1.1_version=1.1.0.csv"
48
+ 3,"ClinGen Hereditary Breast, Ovarian and Pancreatic Cancer Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ATM Version 1",ATM,1.0.0,1/19/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN020?version=1.0.0,"ClinGenHereditaryBreast,OvarianandPancreaticCancerExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforATMVersion1_version=1.0.0.csv"
49
+ 0,ClinGen ACADVL Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,ACADVL,1.0.0,11/9/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN021?version=1.0.0,ClinGenACADVLExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
50
+ 0,ClinGen FBN1 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,FBN1,1.0.0,1/5/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN022?version=1.0.0,ClinGenFBN1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
51
+ 0,ClinGen Hearing Loss Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for OTOF and MYO15A Version 1,"MYO15A, OTOF",1.0.0,3/30/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN023?version=1.0.0,ClinGenHearingLossExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforOTOFandMYO15AVersion1_version=1.0.0.csv
52
+ 0,ClinGen DICER1 and miRNA-Processing Gene Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for DICER1 Version 1.3.0,DICER1,1.3.0,1/30/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN024?version=1.3.0,ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1.3.0_version=1.3.0.csv
53
+ 1,ClinGen DICER1 and miRNA-Processing Gene Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for DICER1 Version 1.2.0,DICER1,1.2.0,5/31/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN024?version=1.2.0,ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1.2.0_version=1.2.0.csv
54
+ 2,ClinGen DICER1 and miRNA-Processing Gene Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for DICER1 Version 1.1.0,DICER1,1.1.0,9/21/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN024?version=1.1.0,ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1.1.0_version=1.1.0.csv
55
+ 3,ClinGen DICER1 and miRNA-Processing Gene Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for DICER1 Version 1,DICER1,1.0.0,5/5/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN024?version=1.0.0,ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1_version=1.0.0.csv
56
+ 0,ClinGen Cerebral Creatine Deficiency Syndromes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GATM Version 2.0.0,GATM,2.0.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN025?version=2.0.0,ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGATMVersion2.0.0_version=2.0.0.csv
57
+ 1,ClinGen Cerebral Creatine Deficiency Syndromes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GATM Version 1.1.0,GATM,1.1.0,9/14/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN025?version=1.1.0,ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGATMVersion1.1.0_version=1.1.0.csv
58
+ 2,ClinGen Cerebral Creatine Deficiency Syndromes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GATM Version 1,GATM,1.0.0,3/21/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN025?version=1.0.0,ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGATMVersion1_version=1.0.0.csv
59
+ 0,ClinGen Cerebral Creatine Deficiency Syndromes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GAMT Version 2.0.0,GAMT,2.0.0,5/23/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN026?version=2.0.0,ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGAMTVersion2.0.0_version=2.0.0.csv
60
+ 1,ClinGen Cerebral Creatine Deficiency Syndromes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GAMT Version 1.1.0,GAMT,1.1.0,9/14/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN026?version=1.1.0,ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGAMTVersion1.1.0_version=1.1.0.csv
61
+ 2,ClinGen Cerebral Creatine Deficiency Syndromes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GAMT Version 1,GAMT,1.0.0,3/21/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN026?version=1.0.0,ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGAMTVersion1_version=1.0.0.csv
62
+ 0,ClinGen Cerebral Creatine Deficiency Syndromes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SLC6A8 Version 1.2.0,SLC6A8,1.2.0,4/10/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN027?version=1.2.0,ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSLC6A8Version1.2.0_version=1.2.0.csv
63
+ 1,ClinGen Cerebral Creatine Deficiency Syndromes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SLC6A8 Version 1.1.0,SLC6A8,1.1.0,9/14/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN027?version=1.1.0,ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSLC6A8Version1.1.0_version=1.1.0.csv
64
+ 2,ClinGen Cerebral Creatine Deficiency Syndromes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SLC6A8 Version 1,SLC6A8,1.0.0,3/21/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN027?version=1.0.0,ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSLC6A8Version1_version=1.0.0.csv
65
+ 0,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TCF4 Version 4.0.0,TCF4,4.0.0,2/28/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN032?version=4.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTCF4Version4.0.0_version=4.0.0.csv
66
+ 1,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TCF4 Version 3.0.0,TCF4,3.0.0,5/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN032?version=3.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTCF4Version3.0.0_version=3.0.0.csv
67
+ 2,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",2.0.0,12/31/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=2.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv
68
+ 3,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",1.0.0,2/17/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=1.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
69
+ 0,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SLC9A6 Version 3.0.0,SLC9A6,3.0.0,5/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN033?version=3.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSLC9A6Version3.0.0_version=3.0.0.csv
70
+ 1,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",2.0.0,12/31/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=2.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv
71
+ 2,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",1.0.0,2/17/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=1.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
72
+ 0,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for CDKL5 Version 4.1.0,CDKL5,4.1.0,3/31/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN034?version=4.1.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforCDKL5Version4.1.0_version=4.1.0.csv
73
+ 1,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for CDKL5 Version 4.0.0,CDKL5,4.0.0,2/26/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN034?version=4.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforCDKL5Version4.0.0_version=4.0.0.csv
74
+ 2,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for CDKL5 Version 3.0.0,CDKL5,3.0.0,5/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN034?version=3.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforCDKL5Version3.0.0_version=3.0.0.csv
75
+ 3,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",2.0.0,12/31/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=2.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv
76
+ 4,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",1.0.0,2/17/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=1.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
77
+ 0,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for FOXG1 Version 4.1.0,FOXG1,4.1.0,3/31/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN035?version=4.1.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforFOXG1Version4.1.0_version=4.1.0.csv
78
+ 1,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for FOXG1 Version 4.0.0,FOXG1,4.0.0,2/26/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN035?version=4.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforFOXG1Version4.0.0_version=4.0.0.csv
79
+ 2,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for FOXG1 Version 3.0.0,FOXG1,3.0.0,5/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN035?version=3.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforFOXG1Version3.0.0_version=3.0.0.csv
80
+ 3,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",2.0.0,12/31/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=2.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv
81
+ 4,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",1.0.0,2/17/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=1.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
82
+ 0,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MECP2 Version 4.1.0,MECP2,4.1.0,3/31/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN036?version=4.1.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMECP2Version4.1.0_version=4.1.0.csv
83
+ 1,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MECP2 Version 3.0.0,MECP2,3.0.0,5/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN036?version=3.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMECP2Version3.0.0_version=3.0.0.csv
84
+ 2,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",2.0.0,12/31/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=2.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv
85
+ 3,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",1.0.0,2/17/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=1.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
86
+ 0,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for UBE3A Version 5.0.0,UBE3A,5.0.0,2/28/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN037?version=5.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforUBE3AVersion5.0.0_version=5.0.0.csv
87
+ 1,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for UBE3A Version 4.0.0,UBE3A,4.0.0,5/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN037?version=4.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforUBE3AVersion4.0.0_version=4.0.0.csv
88
+ 2,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for UBE3A Version 3.0.0,UBE3A,3.0.0,11/10/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN037?version=3.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforUBE3AVersion3.0.0_version=3.0.0.csv
89
+ 3,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 2,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",2.0.0,12/31/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=2.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv
90
+ 4,ClinGen Rett and Angelman-like Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"CDKL5, FOXG1, MECP2, SLC9A6, TCF4, UBE3A",1.0.0,2/17/2021,https://cspec.genome.network/cspec/ui/svi/doc/GN016?version=1.0.0,ClinGenRettandAngelman-likeDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
91
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SHOC2 Version 2.3.0,SHOC2,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN038?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSHOC2Version2.3.0_version=2.3.0.csv
92
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SHOC2 Version 2.2.0,SHOC2,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN038?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSHOC2Version2.2.0_version=2.2.0.csv
93
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SHOC2 Version 2.1.0,SHOC2,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN038?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSHOC2Version2.1.0_version=2.1.0.csv
94
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SHOC2 Version 2.0.0,SHOC2,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN038?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSHOC2Version2.0.0_version=2.0.0.csv
95
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
96
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for NRAS Version 2.3.0,NRAS,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN039?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforNRASVersion2.3.0_version=2.3.0.csv
97
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for NRAS Version 2.2.0,NRAS,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN039?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforNRASVersion2.2.0_version=2.2.0.csv
98
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for NRAS Version 2.1.0,NRAS,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN039?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforNRASVersion2.1.0_version=2.1.0.csv
99
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for NRAS Version 2.0.0,NRAS,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN039?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforNRASVersion2.0.0_version=2.0.0.csv
100
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
101
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RAF1 Version 2.3.0,RAF1,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN040?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRAF1Version2.3.0_version=2.3.0.csv
102
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RAF1 Version 2.2.0,RAF1,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN040?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRAF1Version2.2.0_version=2.2.0.csv
103
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RAF1 Version 2.1.0,RAF1,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN040?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRAF1Version2.1.0_version=2.1.0.csv
104
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RAF1 Version 2.0.0,RAF1,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN040?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRAF1Version2.0.0_version=2.0.0.csv
105
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
106
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SOS1 Version 2.3.0,SOS1,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN041?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSOS1Version2.3.0_version=2.3.0.csv
107
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SOS1 Version 2.2.0,SOS1,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN041?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSOS1Version2.2.0_version=2.2.0.csv
108
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SOS1 Version 2.1.0,SOS1,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN041?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSOS1Version2.1.0_version=2.1.0.csv
109
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SOS1 Version 2.0.0,SOS1,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN041?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSOS1Version2.0.0_version=2.0.0.csv
110
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
111
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SOS2 Version 2.3.0,SOS2,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN042?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSOS2Version2.3.0_version=2.3.0.csv
112
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SOS2 Version 2.2.0,SOS2,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN042?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSOS2Version2.2.0_version=2.2.0.csv
113
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SOS2 Version 2.1.0,SOS2,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN042?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSOS2Version2.1.0_version=2.1.0.csv
114
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SOS2 Version 2.0.0,SOS2,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN042?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSOS2Version2.0.0_version=2.0.0.csv
115
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
116
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PTPN11 Version 2.3.0,PTPN11,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN043?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPTPN11Version2.3.0_version=2.3.0.csv
117
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PTPN11 Version 2.2.0,PTPN11,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN043?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPTPN11Version2.2.0_version=2.2.0.csv
118
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PTPN11 Version 2.1.0,PTPN11,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN043?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPTPN11Version2.1.0_version=2.1.0.csv
119
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PTPN11 Version 2.0.0,PTPN11,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN043?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPTPN11Version2.0.0_version=2.0.0.csv
120
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
121
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for KRAS Version 2.3.0,KRAS,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN044?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforKRASVersion2.3.0_version=2.3.0.csv
122
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for KRAS Version 2.2.0,KRAS,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN044?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforKRASVersion2.2.0_version=2.2.0.csv
123
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for KRAS Version 2.1.0,KRAS,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN044?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforKRASVersion2.1.0_version=2.1.0.csv
124
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for KRAS Version 2.0.0,KRAS,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN044?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforKRASVersion2.0.0_version=2.0.0.csv
125
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
126
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MAP2K1 Version 2.3.0,MAP2K1,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN045?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMAP2K1Version2.3.0_version=2.3.0.csv
127
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MAP2K1 Version 2.2.0,MAP2K1,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN045?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMAP2K1Version2.2.0_version=2.2.0.csv
128
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MAP2K1 Version 2.1.0,MAP2K1,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN045?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMAP2K1Version2.1.0_version=2.1.0.csv
129
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MAP2K1 Version 2.0.0,MAP2K1,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN045?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMAP2K1Version2.0.0_version=2.0.0.csv
130
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
131
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for HRAS Version 2.3.0,HRAS,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN046?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforHRASVersion2.3.0_version=2.3.0.csv
132
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for HRAS Version 2.2.0,HRAS,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN046?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforHRASVersion2.2.0_version=2.2.0.csv
133
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for HRAS Version 2.1.0,HRAS,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN046?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforHRASVersion2.1.0_version=2.1.0.csv
134
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for HRAS Version 2.0.0,HRAS,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN046?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforHRASVersion2.0.0_version=2.0.0.csv
135
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
136
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RIT1 Version 2.3.0,RIT1,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN047?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRIT1Version2.3.0_version=2.3.0.csv
137
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RIT1 Version 2.2.0,RIT1,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN047?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRIT1Version2.2.0_version=2.2.0.csv
138
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RIT1 Version 2.1.0,RIT1,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN047?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRIT1Version2.1.0_version=2.1.0.csv
139
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RIT1 Version 2.0.0,RIT1,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN047?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRIT1Version2.0.0_version=2.0.0.csv
140
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
141
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MAP2K2 Version 2.3.0,MAP2K2,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN048?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMAP2K2Version2.3.0_version=2.3.0.csv
142
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MAP2K2 Version 2.2.0,MAP2K2,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN048?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMAP2K2Version2.2.0_version=2.2.0.csv
143
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MAP2K2 Version 2.1.0,MAP2K2,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN048?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMAP2K2Version2.1.0_version=2.1.0.csv
144
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MAP2K2 Version 2.0.0,MAP2K2,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN048?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMAP2K2Version2.0.0_version=2.0.0.csv
145
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
146
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BRAF Version 2.3.0,BRAF,2.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN049?version=2.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRAFVersion2.3.0_version=2.3.0.csv
147
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BRAF Version 2.2.0,BRAF,2.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN049?version=2.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRAFVersion2.2.0_version=2.2.0.csv
148
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BRAF Version 2.1.0,BRAF,2.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN049?version=2.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRAFVersion2.1.0_version=2.1.0.csv
149
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BRAF Version 2.0.0,BRAF,2.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN049?version=2.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRAFVersion2.0.0_version=2.0.0.csv
150
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines Version 1,"SHOC2, NRAS, RAF1, SOS1, SOS2, PTPN11, KRAS, MAP2K1, HRAS, RIT1, MAP2K2, BRAF",1.0.0,7/18/2017,https://cspec.genome.network/cspec/ui/svi/doc/GN004?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv
151
+ 0,ClinGen Epilepsy Sodium Channel Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SCN1A Version 2.0.0,SCN1A,2.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN067?version=2.0.0,ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1AVersion2.0.0_version=2.0.0.csv
152
+ 1,ClinGen Epilepsy Sodium Channel Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SCN1A Version 1.0.0,SCN1A,1.0.0,3/19/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN067?version=1.0.0,ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1AVersion1.0.0_version=1.0.0.csv
153
+ 0,ClinGen Epilepsy Sodium Channel Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SCN2A Version 2.0.0,SCN2A,2.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN068?version=2.0.0,ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN2AVersion2.0.0_version=2.0.0.csv
154
+ 1,ClinGen Epilepsy Sodium Channel Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SCN2A Version 1.0.0,SCN2A,1.0.0,3/19/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN068?version=1.0.0,ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN2AVersion1.0.0_version=1.0.0.csv
155
+ 0,ClinGen Epilepsy Sodium Channel Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SCN3A Version 2.0.0,SCN3A,2.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN069?version=2.0.0,ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN3AVersion2.0.0_version=2.0.0.csv
156
+ 1,ClinGen Epilepsy Sodium Channel Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SCN3A Version 1.0.0,SCN3A,1.0.0,3/19/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN069?version=1.0.0,ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN3AVersion1.0.0_version=1.0.0.csv
157
+ 0,ClinGen Epilepsy Sodium Channel Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SCN8A Version 2.0.0,SCN8A,2.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN070?version=2.0.0,ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN8AVersion2.0.0_version=2.0.0.csv
158
+ 1,ClinGen Epilepsy Sodium Channel Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SCN8A Version 1.0.0,SCN8A,1.0.0,3/19/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN070?version=1.0.0,ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN8AVersion1.0.0_version=1.0.0.csv
159
+ 0,ClinGen Coagulation Factor Deficiency Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for F8 Version 1.0.0,F8,1.0.0,10/5/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN071?version=1.0.0,ClinGenCoagulationFactorDeficiencyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforF8Version1.0.0_version=1.0.0.csv
160
+ 0,ClinGen Epilepsy Sodium Channel Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SCN1B Version 2.0.0,SCN1B,2.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN076?version=2.0.0,ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1BVersion2.0.0_version=2.0.0.csv
161
+ 1,ClinGen Epilepsy Sodium Channel Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SCN1B Version 1.0.0,SCN1B,1.0.0,3/19/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN076?version=1.0.0,ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1BVersion1.0.0_version=1.0.0.csv
162
+ 0,"ClinGen Hereditary Breast, Ovarian and Pancreatic Cancer Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PALB2 Version 1.1.0",PALB2,1.1.0,11/28/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN077?version=1.1.0,"ClinGenHereditaryBreast,OvarianandPancreaticCancerExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPALB2Version1.1.0_version=1.1.0.csv"
163
+ 1,"ClinGen Hereditary Breast, Ovarian and Pancreatic Cancer Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PALB2 Version 1.0.0",PALB2,1.0.0,3/17/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN077?version=1.0.0,"ClinGenHereditaryBreast,OvarianandPancreaticCancerExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPALB2Version1.0.0_version=1.0.0.csv"
164
+ 0,ClinGen VHL Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for VHL Version 1.1.0,VHL,1.1.0,1/10/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN078?version=1.1.0,ClinGenVHLExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforVHLVersion1.1.0_version=1.1.0.csv
165
+ 1,ClinGen VHL Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for VHL Version 1.0.0,VHL,1.0.0,2/29/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN078?version=1.0.0,ClinGenVHLExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforVHLVersion1.0.0_version=1.0.0.csv
166
+ 0,ClinGen Platelet Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GP1BA Version 1.0.0,GP1BA,1.0.0,2/12/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN079?version=1.0.0,ClinGenPlateletDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGP1BAVersion1.0.0_version=1.0.0.csv
167
+ 0,ClinGen Coagulation Factor Deficiency Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for F9 Version 1.0.0,F9,1.0.0,10/5/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN080?version=1.0.0,ClinGenCoagulationFactorDeficiencyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforF9Version1.0.0_version=1.0.0.csv
168
+ 0,ClinGen von Willebrand Disease Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for VWF Version 1.0.0,VWF,1.0.0,7/9/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN081?version=1.0.0,ClinGenvonWillebrandDiseaseExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforVWFVersion1.0.0_version=1.0.0.csv
169
+ 0,ClinGen Platelet Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GP1BB Version 1.0.0,GP1BB,1.0.0,2/12/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN082?version=1.0.0,ClinGenPlateletDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGP1BBVersion1.0.0_version=1.0.0.csv
170
+ 0,ClinGen Platelet Disorders Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GP9 Version 1.0.0,GP9,1.0.0,2/11/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN083?version=1.0.0,ClinGenPlateletDisordersExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGP9Version1.0.0_version=1.0.0.csv
171
+ 0,ClinGen Thrombosis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SERPINC1 Version 1.1.0,SERPINC1,1.1.0,2/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN084?version=1.1.0,ClinGenThrombosisExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSERPINC1Version1.1.0_version=1.1.0.csv
172
+ 1,ClinGen Thrombosis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SERPINC1 Version 1.0.0,SERPINC1,1.0.0,7/17/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN084?version=1.0.0,ClinGenThrombosisExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSERPINC1Version1.0.0_version=1.0.0.csv
173
+ 0,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for HNF4A Version 2.0.0,HNF4A,2.0.0,10/11/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN085?version=2.0.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforHNF4AVersion2.0.0_version=2.0.0.csv
174
+ 1,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for HNF4A Version 1.1.0,HNF4A,1.1.0,8/11/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN085?version=1.1.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforHNF4AVersion1.1.0_version=1.1.0.csv
175
+ 2,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for HNF4A Version 1.0.0,HNF4A,1.0.0,11/16/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN085?version=1.0.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforHNF4AVersion1.0.0_version=1.0.0.csv
176
+ 0,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GCK Version 2.0.0,GCK,2.0.0,2/17/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN086?version=2.0.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGCKVersion2.0.0_version=2.0.0.csv
177
+ 1,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GCK Version 1.3.0,GCK,1.3.0,8/11/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN086?version=1.3.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGCKVersion1.3.0_version=1.3.0.csv
178
+ 2,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GCK Version 1.2.0,GCK,1.2.0,6/7/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN086?version=1.2.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGCKVersion1.2.0_version=1.2.0.csv
179
+ 3,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GCK Version 1.1.0,GCK,1.1.0,3/23/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN086?version=1.1.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGCKVersion1.1.0_version=1.1.0.csv
180
+ 4,ClinGen Monogenic Diabetes Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GCK Version 1.0.0,GCK,1.0.0,11/16/2022,https://cspec.genome.network/cspec/ui/svi/doc/GN086?version=1.0.0,ClinGenMonogenicDiabetesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGCKVersion1.0.0_version=1.0.0.csv
181
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MRAS Version 1.4.0,MRAS,1.4.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN087?version=1.4.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMRASVersion1.4.0_version=1.4.0.csv
182
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MRAS Version 1.3.0,MRAS,1.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN087?version=1.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMRASVersion1.3.0_version=1.3.0.csv
183
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MRAS Version 1.2.0,MRAS,1.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN087?version=1.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMRASVersion1.2.0_version=1.2.0.csv
184
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MRAS Version 1.1.0,MRAS,1.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN087?version=1.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMRASVersion1.1.0_version=1.1.0.csv
185
+ 4,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MRAS Version 1.0.0,MRAS,1.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN087?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMRASVersion1.0.0_version=1.0.0.csv
186
+ 0,ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for APC Version 2.1.0,APC,2.1.0,11/24/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN089?version=2.1.0,ClinGenInSiGHTHereditaryColorectalCancerPolyposisExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforAPCVersion2.1.0_version=2.1.0.csv
187
+ 1,ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for APC Version 2.0.3,APC,2.0.3,7/24/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN089?version=2.0.3,ClinGenInSiGHTHereditaryColorectalCancerPolyposisExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforAPCVersion2.0.3_version=2.0.3.csv
188
+ 2,ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for APC Version 1.0.0,APC,1.0.0,1/10/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN089?version=1.0.0,ClinGenInSiGHTHereditaryColorectalCancerPolyposisExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforAPCVersion1.0.0_version=1.0.0.csv
189
+ 0,ClinGen von Willebrand Disease Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for VWF Version 1.0.0,VWF,1.0.0,7/9/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN090?version=1.0.0,ClinGenvonWillebrandDiseaseExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforVWFVersion1.0.0_version=1.0.0.csv
190
+ 0,ClinGen Lysosomal Diseases Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for IDUA Version 1.0.0,IDUA,1.0.0,12/5/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN091?version=1.0.0,ClinGenLysosomalDiseasesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforIDUAVersion1.0.0_version=1.0.0.csv
191
+ 0,ClinGen ENIGMA BRCA1 and BRCA2 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BRCA1 Version 1.2.0,BRCA1,1.2.0,1/9/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN092?version=1.2.0,ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA1Version1.2.0_version=1.2.0.csv
192
+ 1,ClinGen ENIGMA BRCA1 and BRCA2 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BRCA1 Version 1.1.0,BRCA1,1.1.0,12/21/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN092?version=1.1.0,ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA1Version1.1.0_version=1.1.0.csv
193
+ 2,ClinGen ENIGMA BRCA1 and BRCA2 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BRCA1 Version 1.0.0,BRCA1,1.0.0,8/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN092?version=1.0.0,ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA1Version1.0.0_version=1.0.0.csv
194
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for LZTR1 Version 1.3.0,LZTR1,1.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN094?version=1.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforLZTR1Version1.3.0_version=1.3.0.csv
195
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for LZTR1 Version 1.2.0,LZTR1,1.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN094?version=1.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforLZTR1Version1.2.0_version=1.2.0.csv
196
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for LZTR1 Version 1.1.0,LZTR1,1.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN094?version=1.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforLZTR1Version1.1.0_version=1.1.0.csv
197
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for LZTR1 Version 1.0.0,LZTR1,1.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN094?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforLZTR1Version1.0.0_version=1.0.0.csv
198
+ 0,ClinGen Cardiomyopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MYBPC3 Version 1.0.0,MYBPC3,1.0.0,4/22/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN095?version=1.0.0,ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYBPC3Version1.0.0_version=1.0.0.csv
199
+ 0,ClinGen ENIGMA BRCA1 and BRCA2 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BRCA2 Version 1.2.0,BRCA2,1.2.0,1/9/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN097?version=1.2.0,ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA2Version1.2.0_version=1.2.0.csv
200
+ 1,ClinGen ENIGMA BRCA1 and BRCA2 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BRCA2 Version 1.1.0,BRCA2,1.1.0,12/21/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN097?version=1.1.0,ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA2Version1.1.0_version=1.1.0.csv
201
+ 2,ClinGen ENIGMA BRCA1 and BRCA2 Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BRCA2 Version 1.0.0,BRCA2,1.0.0,8/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN097?version=1.0.0,ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA2Version1.0.0_version=1.0.0.csv
202
+ 0,ClinGen Cardiomyopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TNNI3 Version 1.0.0,TNNI3,1.0.0,4/22/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN098?version=1.0.0,ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTNNI3Version1.0.0_version=1.0.0.csv
203
+ 0,ClinGen Cardiomyopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TNNT2 Version 1.0.0,TNNT2,1.0.0,4/22/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN099?version=1.0.0,ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTNNT2Version1.0.0_version=1.0.0.csv
204
+ 0,ClinGen Cardiomyopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for TPM1 Version 1.0.0,TPM1,1.0.0,4/22/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN100?version=1.0.0,ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTPM1Version1.0.0_version=1.0.0.csv
205
+ 0,ClinGen Cardiomyopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ACTC1 Version 1.0.0,ACTC1,1.0.0,4/22/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN101?version=1.0.0,ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACTC1Version1.0.0_version=1.0.0.csv
206
+ 0,ClinGen Cardiomyopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MYL2 Version 1.0.0,MYL2,1.0.0,4/22/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN102?version=1.0.0,ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYL2Version1.0.0_version=1.0.0.csv
207
+ 0,ClinGen Cardiomyopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MYL3 Version 1.0.0,MYL3,1.0.0,4/22/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN103?version=1.0.0,ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYL3Version1.0.0_version=1.0.0.csv
208
+ 0,ClinGen Severe Combined Immunodeficiency Disease Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for FOXN1 Version 1.0.0,FOXN1,1.0.0,7/29/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN113?version=1.0.0,ClinGenSevereCombinedImmunodeficiencyDiseaseExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforFOXN1Version1.0.0_version=1.0.0.csv
209
+ 0,ClinGen Severe Combined Immunodeficiency Disease Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ADA Version 1.0.0,ADA,1.0.0,10/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN114?version=1.0.0,ClinGenSevereCombinedImmunodeficiencyDiseaseExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforADAVersion1.0.0_version=1.0.0.csv
210
+ 0,ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MLH1 Version 1.0.0,MLH1,1.0.0,8/9/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN115?version=1.0.0,ClinGenInSiGHTHereditaryColorectalCancerPolyposisExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMLH1Version1.0.0_version=1.0.0.csv
211
+ 0,ClinGen Severe Combined Immunodeficiency Disease Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for DCLRE1C Version 1.0.0,DCLRE1C,1.0.0,10/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN116?version=1.0.0,ClinGenSevereCombinedImmunodeficiencyDiseaseExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDCLRE1CVersion1.0.0_version=1.0.0.csv
212
+ 0,ClinGen Severe Combined Immunodeficiency Disease Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for IL7R Version 1.0.0,IL7R,1.0.0,10/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN119?version=1.0.0,ClinGenSevereCombinedImmunodeficiencyDiseaseExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforIL7RVersion1.0.0_version=1.0.0.csv
213
+ 0,ClinGen Leber Congenital Amaurosis/early onset Retinal Dystrophy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RPE65 Version 1.0.0,RPE65,1.0.0,10/24/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN120?version=1.0.0,ClinGenLeberCongenitalAmaurosisearlyonsetRetinalDystrophyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRPE65Version1.0.0_version=1.0.0.csv
214
+ 0,ClinGen Severe Combined Immunodeficiency Disease Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for JAK3 Version 1.0.0,JAK3,1.0.0,10/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN121?version=1.0.0,ClinGenSevereCombinedImmunodeficiencyDiseaseExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforJAK3Version1.0.0_version=1.0.0.csv
215
+ 0,ClinGen Severe Combined Immunodeficiency Disease Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RAG1 Version 1.0.0,RAG1,1.0.0,10/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN123?version=1.0.0,ClinGenSevereCombinedImmunodeficiencyDiseaseExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRAG1Version1.0.0_version=1.0.0.csv
216
+ 0,ClinGen Severe Combined Immunodeficiency Disease Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RAG2 Version 1.0.0,RAG2,1.0.0,10/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN124?version=1.0.0,ClinGenSevereCombinedImmunodeficiencyDiseaseExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRAG2Version1.0.0_version=1.0.0.csv
217
+ 0,ClinGen Pulmonary Hypertension Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BMPR2 Version 1.1.0,BMPR2,1.1.0,4/6/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN125?version=1.1.0,ClinGenPulmonaryHypertensionExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBMPR2Version1.1.0_version=1.1.0.csv
218
+ 1,ClinGen Pulmonary Hypertension Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for BMPR2 Version 1.0.0,BMPR2,1.0.0,3/1/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN125?version=1.0.0,ClinGenPulmonaryHypertensionExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBMPR2Version1.0.0_version=1.0.0.csv
219
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RRAS2 Version 1.3.0,RRAS2,1.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN127?version=1.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRRAS2Version1.3.0_version=1.3.0.csv
220
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RRAS2 Version 1.2.0,RRAS2,1.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN127?version=1.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRRAS2Version1.2.0_version=1.2.0.csv
221
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RRAS2 Version 1.1.0,RRAS2,1.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN127?version=1.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRRAS2Version1.1.0_version=1.1.0.csv
222
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RRAS2 Version 1.0.0,RRAS2,1.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN127?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRRAS2Version1.0.0_version=1.0.0.csv
223
+ 0,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PPP1CB Version 1.3.0,PPP1CB,1.3.0,12/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN128?version=1.3.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPPP1CBVersion1.3.0_version=1.3.0.csv
224
+ 1,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PPP1CB Version 1.2.0,PPP1CB,1.2.0,12/2/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN128?version=1.2.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPPP1CBVersion1.2.0_version=1.2.0.csv
225
+ 2,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PPP1CB Version 1.1.0,PPP1CB,1.1.0,9/17/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN128?version=1.1.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPPP1CBVersion1.1.0_version=1.1.0.csv
226
+ 3,ClinGen RASopathy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PPP1CB Version 1.0.0,PPP1CB,1.0.0,8/3/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN128?version=1.0.0,ClinGenRASopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPPP1CBVersion1.0.0_version=1.0.0.csv
227
+ 0,ClinGen Severe Combined Immunodeficiency Disease Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for IL2RG Version 1.0.0,IL2RG,1.0.0,10/9/2023,https://cspec.genome.network/cspec/ui/svi/doc/GN129?version=1.0.0,ClinGenSevereCombinedImmunodeficiencyDiseaseExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforIL2RGVersion1.0.0_version=1.0.0.csv
228
+ 0,ClinGen Hereditary Hemorrhagic Telangiectasia Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ACVRL1 Version 1.1.0,ACVRL1,1.1.0,3/20/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN135?version=1.1.0,ClinGenHereditaryHemorrhagicTelangiectasiaExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACVRL1Version1.1.0_version=1.1.0.csv
229
+ 1,ClinGen Hereditary Hemorrhagic Telangiectasia Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ACVRL1 Version 1.0.0,ACVRL1,1.0.0,3/5/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN135?version=1.0.0,ClinGenHereditaryHemorrhagicTelangiectasiaExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACVRL1Version1.0.0_version=1.0.0.csv
230
+ 0,ClinGen Hereditary Hemorrhagic Telangiectasia Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ENG Version 1.1.0,ENG,1.1.0,3/20/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN136?version=1.1.0,ClinGenHereditaryHemorrhagicTelangiectasiaExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforENGVersion1.1.0_version=1.1.0.csv
231
+ 1,ClinGen Hereditary Hemorrhagic Telangiectasia Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ENG Version 1.0.0,ENG,1.0.0,3/5/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN136?version=1.0.0,ClinGenHereditaryHemorrhagicTelangiectasiaExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforENGVersion1.0.0_version=1.0.0.csv
232
+ 0,ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MSH2 Version 1.0.0,MSH2,1.0.0,8/9/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN137?version=1.0.0,ClinGenInSiGHTHereditaryColorectalCancerPolyposisExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMSH2Version1.0.0_version=1.0.0.csv
233
+ 0,ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MSH6 Version 1.0.0,MSH6,1.0.0,8/9/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN138?version=1.0.0,ClinGenInSiGHTHereditaryColorectalCancerPolyposisExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMSH6Version1.0.0_version=1.0.0.csv
234
+ 0,ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PMS2 Version 1.0.0,PMS2,1.0.0,8/9/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN139?version=1.0.0,ClinGenInSiGHTHereditaryColorectalCancerPolyposisExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPMS2Version1.0.0_version=1.0.0.csv
235
+ 1,ClinGen InSiGHT Hereditary Colorectal Cancer/Polyposis Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for PMS2 Version 1.0.0,PMS2,1.0.0,8/9/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN139?version=1.0.0,ClinGenInSiGHTHereditaryColorectalCancerPolyposisExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforPMS2Version1.0.0_version=1.0.0.csv
236
+ 0,ClinGen Congenital Myopathies Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for NEB Version 1.0.0,NEB,1.0.0,8/7/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN146?version=1.0.0,ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforNEBVersion1.0.0_version=1.0.0.csv
237
+ 0,ClinGen Congenital Myopathies Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ACTA1 Version 2.0.0,ACTA1,2.0.0,8/27/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN147?version=2.0.0,ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACTA1Version2.0.0_version=2.0.0.csv
238
+ 1,ClinGen Congenital Myopathies Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ACTA1 Version 1.0.0,ACTA1,1.0.0,8/7/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN147?version=1.0.0,ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACTA1Version1.0.0_version=1.0.0.csv
239
+ 0,ClinGen Congenital Myopathies Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for DNM2 Version 1.0.0,DNM2,1.0.0,8/7/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN148?version=1.0.0,ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDNM2Version1.0.0_version=1.0.0.csv
240
+ 0,ClinGen Congenital Myopathies Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for MTM1 Version 1.0.0,MTM1,1.0.0,8/7/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN149?version=1.0.0,ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMTM1Version1.0.0_version=1.0.0.csv
241
+ 0,ClinGen Congenital Myopathies Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RYR1 Version 2.0.0,RYR1,2.0.0,12/12/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN150?version=2.0.0,ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRYR1Version2.0.0_version=2.0.0.csv
242
+ 1,ClinGen Congenital Myopathies Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RYR1 Version 1.0.0,RYR1,1.0.0,8/7/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN150?version=1.0.0,ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRYR1Version1.0.0_version=1.0.0.csv
243
+ 0,ClinGen Leber Congenital Amaurosis/early onset Retinal Dystrophy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for GUCY2D Version 1.0.0,GUCY2D,1.0.0,1/22/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN167?version=1.0.0,ClinGenLeberCongenitalAmaurosisearlyonsetRetinalDystrophyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGUCY2DVersion1.0.0_version=1.0.0.csv
244
+ 0,ClinGen Congenital Myopathies Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ACTA1 Version 1.0.0,ACTA1,1.0.0,8/7/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN169?version=1.0.0,ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACTA1Version1.0.0_version=1.0.0.csv
245
+ 0,ClinGen Congenital Myopathies Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RYR1 Version 2.0.0,RYR1,2.0.0,12/12/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN179?version=2.0.0,ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRYR1Version2.0.0_version=2.0.0.csv
246
+ 1,ClinGen Congenital Myopathies Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for RYR1 Version 1.0.0,RYR1,1.0.0,8/7/2024,https://cspec.genome.network/cspec/ui/svi/doc/GN179?version=1.0.0,ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRYR1Version1.0.0_version=1.0.0.csv
247
+ 0,ClinGen Limb Girdle Muscular Dystrophy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for DYSF Version 1.0.0,DYSF,1.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN180?version=1.0.0,ClinGenLimbGirdleMuscularDystrophyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDYSFVersion1.0.0_version=1.0.0.csv
248
+ 0,ClinGen Limb Girdle Muscular Dystrophy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SGCB Version 1.0.0,SGCB,1.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN184?version=1.0.0,ClinGenLimbGirdleMuscularDystrophyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSGCBVersion1.0.0_version=1.0.0.csv
249
+ 0,ClinGen Limb Girdle Muscular Dystrophy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SGCG Version 1.0.0,SGCG,1.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN185?version=1.0.0,ClinGenLimbGirdleMuscularDystrophyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSGCGVersion1.0.0_version=1.0.0.csv
250
+ 0,ClinGen Limb Girdle Muscular Dystrophy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SGCD Version 1.0.0,SGCD,1.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN186?version=1.0.0,ClinGenLimbGirdleMuscularDystrophyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSGCDVersion1.0.0_version=1.0.0.csv
251
+ 0,ClinGen Limb Girdle Muscular Dystrophy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for CAPN3 Version 1.0.0,CAPN3,1.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN187?version=1.0.0,ClinGenLimbGirdleMuscularDystrophyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforCAPN3Version1.0.0_version=1.0.0.csv
252
+ 0,ClinGen Limb Girdle Muscular Dystrophy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for ANO5 Version 1.0.0,ANO5,1.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN188?version=1.0.0,ClinGenLimbGirdleMuscularDystrophyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforANO5Version1.0.0_version=1.0.0.csv
253
+ 0,ClinGen Limb Girdle Muscular Dystrophy Expert Panel Specifications to the ACMG/AMP Variant Interpretation Guidelines for SGCA Version 1.0.0,SGCA,1.0.0,1/7/2025,https://cspec.genome.network/cspec/ui/svi/doc/GN189?version=1.0.0,ClinGenLimbGirdleMuscularDystrophyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSGCAVersion1.0.0_version=1.0.0.csv
VCI/parsing_csr_criteria/cspec_version_guide_processed.csv ADDED
The diff for this file is too large to render. See raw diff
 
VCI/parsing_csr_criteria/version_csv_individual/ClinGenACADVLExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv ADDED
@@ -0,0 +1,255 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ ACADVL (HGNC:92),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ ACADVL (HGNC:92),PVS1,Very Strong,"Loss of function is a known mechanism for VLCAD Deficiency. The specifications below are based on published guidance for assigning strength of evidence for PVS1 (Abou Tayoun et al., 2018; PMID: 30192042). There are multiple transcripts for ACADVL. The major isoform, NM_000018.4, encodes a 655 amino acid precursor protein that contains a 40 amino acid N-terminal target sequence that is removed during uptake (Aoyama et al., 1995; PMID: 7668252). In a joint project between NCBI and EMBL-EBI (MANE), NM_000018.4 was designated as the most relevant transcript.
9
+ Nonsense or Frameshift:
10
+
11
+
12
+
13
+
14
+ Use caution when interpreting LOF variants at the 3’ end of the gene.
15
+
16
+
17
+ All nonsense and frameshift variants will meet PVS1 unless the variant is predicted to be missed by nonsense-mediated decay (NMD).
18
+
19
+
20
+ NMD is not predicted if the variant is in the last exon (exon 20) or in the last 50 nucleotides of the penultimate exon (exon 19).
21
+ Canonical Splice Site (+1, +2, -1, -2):
22
+
23
+
24
+ All donor/acceptor sites follow the GT/AG rule, except for the donor splice site of intron 8, which begins with GC. PVS1 should not be applied for variants in the splice donor site of intron 8 since the impact of GC donor splice sites is not well understood.
25
+
26
+
27
+ For +1 or +2 GT donor splice site variants, the exon immediately 5’ of the variant is predicted to be skipped. For -1 or -2 AG acceptor splice site variants, the exon immediately 3’ of the variants is predicted to be skipped. For the predicted in frame/out of frame consequences for exon skipping in ACADVL (NM_000018.4), see Appendix 1.
28
+ Deletions:
29
+
30
+
31
+ For single or multi-exon deletions that result in an out-of-frame consequence, use PVS1 unless NMD is not predicted to occur. If NMD is not predicted to occur, use PVS1_Moderate.
32
+
33
+
34
+ If a deletion results in an in-frame consequence, the deletion must encompass one or more exons in order to apply PVS1. Consult Appendix 1 and the PVS1 decision tree to assign a strength.
35
+ Duplications:
36
+
37
+
38
+ Single and multi-exon duplications have not been reported in ACADVL. Consult the PVS1 decision tree to assign the strength.
39
+ Initiation codon:
40
+
41
+
42
+ The next in-frame methionine is at position 6 (on transcript NM_000018). However, the first 40 amino acids comprise the leader sequence in the precursor peptide and are important for proper localization of the protein (Aoyama et al., 1995; PMID: 7668252). Therefore, initiator codon variants will meet PVS1_Strong.",Disease-specific
43
+ ACADVL (HGNC:92),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
44
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
45
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
46
+ ACADVL (HGNC:92),PS1,Strong,Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.,No change
47
+ ACADVL (HGNC:92),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
48
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
49
+ ACADVL (HGNC:92),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
50
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
51
+ ACADVL (HGNC:92),PS3,Strong,"Functional evidence from non-patient derived material with only a single variant best reflects the variant-level phenotype. Apply patient-derived evidence in PP4.
52
+
53
+
54
+ Apply criteria at the level determined by validation parameters (see flowchart below). VLCAD assays available now do not meet the criteria that is being proposed now regarding the types of controls etc. but the published VLCAD assays are well established with many positive and negative controls being run.
55
+
56
+
57
+ Hesse et al., 2018 (PMID 30194637) reviewed enzymatic testing in lymphocytes as a confirmatory tool in newborns identified by screening. Molecular testing was performed after in most patients.
58
+
59
+
60
+ If an enzyme activity assay has >20% activity it cannot be weighted above PS3_supporting regardless of flowchart results.",Disease-specific
61
+ ACADVL (HGNC:92),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
62
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
63
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",NA
64
+ ACADVL (HGNC:92),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
65
+ ACADVL (HGNC:92),PM1,Moderate,"Examples of mutational hot spots:
66
+ It has been cited that the CpG dinucleotides in the codon for arginine326 and arginine429 are mutational hot spots, since CrT (R326C and R429W) and GrA (R326H and R429Q) mutations of the CpG di- nucleotide are present in both codons. Article for reference is Clear Correlation of Genotype with Disease Phenotype in Very–Long-Chain Acyl-CoA Dehydrogenase Deficiency – by Andresen et al., 1999 (PMID 9973285).
67
+
68
+
69
+ Please refer to “Structural Basis for Substrate Fatty Acyl Chain Specificity: Crystal Structure of Human Very-Long-Chain acyl-CoA Dehydrogenase” by McAndrew et al., 2008 (PMID 18227065) (note – numbering in this reference is for the mature peptide without the mitochondrial signal peptide, so amino acid position numbers are lower by 40 than numbering based on NM-000018.4) and “Compared effects of missense mutations in Very-Long-Chain Acyl-CoA Dehydrogenase deficiency: Combined analysis by structural, functional and pharmacological approaches” by Gobin-Limballe et al., 2010 (PMID 20060901) for identifying important structural regions for proper enzyme function such as binding sites and active sites of the enzyme. Based on this and information from UniProt the following regions are designated to be important functional domains:
70
+
71
+
72
+ p.214-223, p.249-251, p.460-466, p.562: Nucleotide and substrate binding.
73
+
74
+
75
+ p.481-516: Membrane binding.
76
+
77
+
78
+ p.1-40: Mitochondrial signal peptide.
79
+
80
+
81
+ Curators may seek approval from the expert panel for identifying new hot spots or critical regions as discovered in literature searches, in uniport, HGMD, etc. for inclusion.",Disease-specific
82
+ ACADVL (HGNC:92),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
83
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
84
+ ACADVL (HGNC:92),PM2,Supporting,"Variants with a highest population minor allele frequency (MAF) <0.001 (0.1%) in any continental population with >2000 alleles in gnomAD will meet PM2_supporting.
85
+
86
+
87
+ The 0.001 cutoff is based on a disease frequency of 1:100,000, genetic heterogeneity of 100%, penetrance of 75%, and maximum allelic contribution of 20%, based on the most common pathogenic ACADVL variant, c.848T>C (p.Val283Ala): gnomAD MAF: 0.002238 (289/129106; 1 homozygote; European Non-Finnish) and overall frequency: 0.001224 (346/282744; 2 homozygotes). This variant would not meet PM2_supporting, but it is assumed that the most common variants have already been identified so a conservative cutoff was chosen.
88
+
89
+
90
+ See Appendix 2 for calculations.
91
+
92
+
93
+ It is acceptable for an ACADVL variant to be present in controls because VLCAD deficiency is a recessive condition. It is also possible for homozygous ACADVL variants to be present in population databases due to later onset of the condition. If homozygous variants are present, the number should be noted and discussed with an expert.
94
+
95
+
96
+ SVI guidance: use PM2 as Supporting (not Moderate) based on absence or rarity not being in the odds range expected for a moderate piece of evidence. This EP will track the impact of not using Moderate for PM2.",Disease-specific
97
+ ACADVL (HGNC:92),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
98
+ Note: This requires testing of parents (or offspring) to determine phase.",
99
+ ACADVL (HGNC:92),PM3,Moderate,"Details of the cDNA change must be used to describe any variants used as evidence for PM3. If the variant is described only as an amino acid change, this is not sufficient. Probands must also meet PP4 criteria to be counted.
100
+
101
+
102
+ If more than one case has the same genotype and the variants are not confirmed in trans, then only one case should be used for assigning points to avoid overcounting evidence if the variants are actually in cis and hence inherited together in multiple individuals or potentially counting the same case twice. If the variants are confirmed to be in trans, more than one individual with the same genotype can be counted as long as the reports do not represent the same case.
103
+
104
+
105
+ Following SVI guidance for PM3, use the scoring system below to determine the strength of evidence for PM3.
106
+
107
+
108
+ These variant interpretation guidelines should be used to determine the classification of the “other variant” in order to determine the appropriate number of points to assign.
109
+
110
+
111
+ For a variant to be “confirmed in trans” in a compound heterozygous patient, parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed. Otherwise, the phase of the variants is unknown. Parental testing is not required for homozygous cases.",Disease-specific
112
+ ACADVL (HGNC:92),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
113
+ ACADVL (HGNC:92),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
114
+ ACADVL (HGNC:92),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
115
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
116
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
117
+ ACADVL (HGNC:92),PM5,Moderate,"These variant interpretation guidelines should be used to determine the classification of the other missense change in order to determine whether this rule can be applied.
118
+
119
+
120
+ Use PM5_Supporting if the other variant is Likely Pathogenic.",Disease-specific
121
+ ACADVL (HGNC:92),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
122
+ ACADVL (HGNC:92),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity.",No change
123
+ ACADVL (HGNC:92),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
124
+ Note: May be used as stronger evidence with increasing segregation data.",
125
+ ACADVL (HGNC:92),PP1,Supporting,"Following SVI guidance, use the scoring system below to determine the strength of evidence for PP1.
126
+
127
+
128
+ For segregation counting, do not count probands as a segregation.
129
+ o Affected segregations = # affected individuals in the family with the variants - 1.
130
+
131
+
132
+ Affected segregations are defined as affected family members (typically siblings) who harbor the variant in question and a second variant on the remaining allele.
133
+
134
+
135
+ Unaffected segregations are defined as unaffected family members, typically siblings, who are at risk to inherit the two variants identified in the proband. These individuals should be either wild-type for both variants identified in the proband, or a heterozygous carrier for a single variant.
136
+
137
+
138
+ Unaffected, carrier parents DO NOT count as unaffected segregations.
139
+
140
+
141
+ There may be scenarios where individuals other than siblings could be counted as segregations, such as in families where one parent is affected with the autosomal recessive disorder, in large families with multiple branches, or in consanguineous families.",Disease-specific
142
+ ACADVL (HGNC:92),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
143
+ ACADVL (HGNC:92),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
144
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
145
+ ACADVL (HGNC:92),PP3,Supporting,"Missense changes with a REVEL score >0.75 will meet PP3.
146
+
147
+
148
+ For in-frame deletions and insertions, use PROVEAN and Mutation Taster. Results must be consistent to count.
149
+
150
+
151
+ For non-canonical splice site variants, use Splice AI, MaxEntScn and NNSplice. Based on data from Jaganathan et al., 2019 (PMID: 30661751), Houdayer et al., 2012 (PMID: 22505045) and Tang et al., 2016 (PMID: 27313609), PP3 can be applied if there is, a SpliceAI “high score” (Δ Score ≥ 0.5 “confidently predicted splice variants”) (exclude any results with Δ Score ≤ 0.2 from consideration of pathogenicity, <0.2 are not “predicted to alter splicing”), >15% reduction using MaxEntScn and >5% reduction using NNSplice. Two out of the three predictors must be consistent to count.
152
+
153
+
154
+ For SpliceAI’s cryptic splice-site rules, the creation of a new splice-site with Δ Score ≥ 0.5 may be enough to produce a large proportion of aberrant transcripts.
155
+
156
+
157
+ If a new splice site is predicted to be generated, this rule can be applied if the newly generated splice site is significantly stronger than the wild type site (Δ Score ≥ 0.5 using SpliceAI; >15% difference using MaxEntScn).
158
+
159
+
160
+ Do not apply this rule for canonical splice site changes meeting PVS1.",Disease-specific
161
+ ACADVL (HGNC:92),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
162
+ ACADVL (HGNC:92),PP4,Moderate,"Abnormal tests that are consistent with VLCAD deficiency include deficient VLCAD enzyme activity in patient cells (leukocytes, fibroblasts, liver, heart, or skeletal muscle, or amniocytes), abnormal C14:1 acylcarnitine values from newborn screening (NBS), and abnormal acylcarnitine values from follow-up plasma analysis.
163
+
164
+
165
+ To award a given tier of evidence, at least one proband harboring the variant must meet at least one of the minimum requirements below. If multiple analyses are completed, only one result needs to be within the allowed ranges to meet this rule as values can fluctuate due to age or diet:
166
+
167
+
168
+ PP4_moderate (Must meet at least one of the below criteria):
169
+
170
+
171
+ VLCAD enzyme activity (β-Oxidation Flux) ≤20% of normal.
172
+
173
+
174
+ NBS C14:1 Levels ≥ 1.0 μM AND:
175
+
176
+
177
+ VLCAD enzyme activity (β-Oxidation Flux) 21-27% of normal OR Assertion of reduced VLCAD activity without specific levels OR Follow-Up Plasma Acylcarnitine analysis “consistent with VLCADD” without specific levels.",Disease-specific
178
+ ACADVL (HGNC:92),PP4,Supporting,"Abnormal tests that are consistent with VLCAD deficiency include deficient VLCAD enzyme activity in patient cells (leukocytes, fibroblasts, liver, heart, or skeletal muscle, or amniocytes), abnormal C14:1 acylcarnitine values from newborn screening (NBS), and abnormal acylcarnitine values from follow-up plasma analysis.
179
+
180
+
181
+ To award a given tier of evidence, at least one proband harboring the variant must meet at least one of the minimum requirements below. If multiple analyses are completed, only one result needs to be within the allowed ranges to meet this rule as values can fluctuate due to age or diet:
182
+
183
+
184
+ PP4_supporting (Must meet at least one of the below criteria):
185
+
186
+
187
+ VLCAD enzyme activity (β-Oxidation Flux) 21-27% of normal.
188
+
189
+
190
+ Assertion of reduced VLCAD activity without specific levels.
191
+
192
+
193
+ NBS C14:1 Levels from >0.8 μM.
194
+
195
+
196
+ Assertion of abnormal NBS “consistent with VLCADD” without specific levels AND: Follow-Up Plasma Acylcarnitine analysis “consistent with VLCADD” without specific levels.",Disease-specific
197
+ ACADVL (HGNC:92),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
198
+ ACADVL (HGNC:92),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
199
+ ACADVL (HGNC:92),BA1,Stand Alone,"Variants with a highest population minor allele frequency (MAF) ≥0.007 (0.7%) in any continental population with >2000 alleles in gnomAD will meet BA1.
200
+
201
+
202
+ See Appendix 2 for calculations.",Disease-specific
203
+ ACADVL (HGNC:92),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
204
+ ACADVL (HGNC:92),BS1,Strong,"Variants with a highest population minor allele frequency (MAF) ≥0.0035 (0.35%) in any continental population with >2000 alleles in gnomAD will meet BS1.
205
+
206
+
207
+ See Appendix 2 for calculations.",Disease-specific
208
+ ACADVL (HGNC:92),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",NA
209
+ ACADVL (HGNC:92),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
210
+ ACADVL (HGNC:92),BS3,Strong,"In vitro expression:
211
+
212
+
213
+
214
+
215
+
216
+
217
+ 50% enzyme activity
218
+ Splicing assays:
219
+
220
+
221
+
222
+
223
+
224
+
225
+ For non-canonical splicing variants, use BS3 if there is evidence demonstrating normal splicing with no evidence of abnormal splicing (RT-PCR and/or RNA sequencing).
226
+
227
+
228
+ These studies can be performed using patient-derived cells or heterologous cultured cells.
229
+
230
+
231
+ Any variants meeting the requirement for in vitro expression or splicing assays can meet BS3, but if the variant meets the description for both, BS3 should only be counted once.",Disease-specific
232
+ ACADVL (HGNC:92),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
233
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
234
+ ACADVL (HGNC:92),BS4,Strong,Lack of segregation in affected members of a family.,No change
235
+ ACADVL (HGNC:92),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
236
+ ACADVL (HGNC:92),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
237
+ ACADVL (HGNC:92),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder; or observed in cis with a pathogenic variant in any inheritance pattern.,No change
238
+ ACADVL (HGNC:92),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
239
+ ACADVL (HGNC:92),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
240
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
241
+ ACADVL (HGNC:92),BP4,Supporting,"Missense changes with a REVEL score <0.5 will meet BP4.
242
+
243
+
244
+ For in-frame deletions and insertions, use PROVEAN and Mutation Taster. Results must be consistent to count.
245
+ For non-canonical splice site variants, use Splice AI, MaxEntScn and NNSplice. Based on data from Jaganathan et al., 2019 (PMID: 30661751), Houdayer et al., 2012 (PMID: 22505045) and Tang et al., 2016 (PMID: 27313609), PP3 can be applied if there is a with Δ Score ≤ 0.2 , <10% reduction using MaxEntScn and <2% reduction using NNSplice. Two out of the three predictors must be consistent to count.
246
+
247
+
248
+ Do not apply this rule if there is evidence for creation of a cryptic splice site.
249
+
250
+
251
+ Can be used with BP7 code.",Disease-specific
252
+ ACADVL (HGNC:92),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
253
+ ACADVL (HGNC:92),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
254
+ ACADVL (HGNC:92),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
255
+ ACADVL (HGNC:92),BP7,Supporting,Can be used with BP4 code.,No change
VCI/parsing_csr_criteria/version_csv_individual/ClinGenBrainMalformationsExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1.1.0_version=1.1.0.csv ADDED
@@ -0,0 +1,104 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ AKT3 (HGNC:393),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",NA
8
+ AKT3 (HGNC:393),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
9
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
10
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
11
+ AKT3 (HGNC:393),PS1,Strong,No change.,None
12
+ AKT3 (HGNC:393),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
13
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
14
+ AKT3 (HGNC:393),PS2,Strong,"Award the PS2_Strong point if Criteria 1 AND Criteria 2 are fulfilled.
15
+
16
+
17
+ Criteria 1. The variant is present at a detectable allele fraction but is absent from parental samples with confirmed maternity and paternity.
18
+
19
+
20
+ Criteria 2. The variant is present at a detectable allele fraction in an affected tissue sample but is absent from or detected at a lower allelic fraction in another tissue (e.g. if present in 5% of brain tissue but absent from the blood or skin this point can be awarded).
21
+
22
+
23
+ For the sake of implementation, these criteria are intended to apply to high-confidence somatic mutations identified by the reporting CLIA laboratory. The expert panel recognizes that in practice there may be significant heterogeneity in the technical methods and thresholds used to identify such variants as 'high confidence', and flags the need to establish consensus statistical frameworks (e.g. Phred-scaled genotype qualities) or experimental approaches (e.g., confirmation of somatic variants by sequencing on orthogonal platforms) by which quality thresholds can be consistently applied.","Disease-specific,Strength"
24
+ AKT3 (HGNC:393),PS2,Moderate,"Award the PS2_Moderate point if Criteria 1 is fulfilled, OR if parents are not available but Criteria 2 is fulfilled.","Disease-specific,Strength"
25
+ AKT3 (HGNC:393),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
26
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
27
+ AKT3 (HGNC:393),PS3,Strong,"Follow recommendations set forth by the SVI in conjunction with specifications added by the BMVCEP for quality metrics and minimum validation controls required. (Supplemental Document 1) Animal models are considered in a different manner.
28
+
29
+
30
+ Award PS4_Strong if the animal model generated with the variant of interest expressed in neural progenitors shows a complementary brain phenotype.
31
+
32
+
33
+ Award PS3 if the functional assay meets the acceptability criteria delimited in (PMID: 31892348) with specifications added by the BMVCEP. Quality metrics and minimum validation controls required can be found in Supplementary Document 1.
34
+
35
+
36
+ Animal models are considered in a different manner. Award PS4_Strong if the animal model generated with the variant of interest expressed in neural progenitors show a complementary brain phenotype.
37
+
38
+
39
+ Caveat: Studies of cell lines derived from the affected patient as the only source of functional characterization are by themselves insufficient to provide strong evidence of pathogenicity. This is because cells derived from patient affected tissue are likely to exhibit the desired phenotype since the patient tissue exhibits the phenotype. It is therefore impossible to determine whether the variant of interest was solely responsible for that phenotype. Instead, functional readout of patient-derived cells are now included in PS4.",Disease-specific
40
+ AKT3 (HGNC:393),PS3,Moderate,Follow recommendations set forth by the SVI in conjunction with specifications added by the BMVCEP for quality metrics and minimum validation controls required (PMID: 31892348). Animal models are considered in a different manner. Award PS4_Moderate if the animal model generated with the variant of interest expressed in non-neural tissues show an increased cancer burden.,Strength
41
+ AKT3 (HGNC:393),PS3,Supporting,Follow recommendations set forth by the SVI in conjunction with specifications added by the BMVCEP for quality metrics and minimum validation controls required (PMID: 31892348).,Strength
42
+ AKT3 (HGNC:393),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
43
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
44
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
45
+ AKT3 (HGNC:393),PS4,Very Strong,"Points are assigned for phenotype according to (Table 2A). Phenotype criteria can only be used if the variant meets criteria for (PM2). Strength of evidence is determined by points according to (Table 2B). PS4_VeryStrong ≥ 16 points. For PS4, for cases reported in the literature, we recommend assigning each one to the SINGLE category below that is associated with the highest point value (Table 2A). The total score obtained for all reported cases with a particular variant will determine the strength of PS4 assigned according to the scale (Table 2B)
46
+ *
47
+ .
48
+
49
+
50
+ PS4_VeryStrong ≥ 16 points.
51
+
52
+
53
+ *
54
+ Applicable if the variant is absent/rare from controls according to PM2 to ensure the variant is not simply present due to beinging common in the general population.","Disease-specific,Strength"
55
+ AKT3 (HGNC:393),PS4,Strong,Points are assigned for phenotype according to (Table 2A). Phenotype criteria can only be used if the variant is absent from controls (PM2). Strength of evidence is determined by points according to (Table 2B). PS4_Strong = 3.5-15.75 points.,Disease-specific
56
+ AKT3 (HGNC:393),PS4,Moderate,Points are assigned for phenotype according to (Table 2A). Phenotype criteria can only be used if the variant is absent from controls (PM2). Strength of evidence is determined by points according to (Table 2B). PS4_Moderate = 1.5-3.25 points.,Strength
57
+ AKT3 (HGNC:393),PS4,Supporting,Points are assigned for phenotype according to (Table 2A). Phenotype criteria can only be used if the variant is absent from controls (PM2). Strength of evidence is determined by points according to (Table 2B). PS4_Supporting = 0.5 – 1.25 points.,Strength
58
+ AKT3 (HGNC:393),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
59
+ AKT3 (HGNC:393),PM1,Supporting,Residues affecting critical functional domains provided in Table 4 for each gene.,Strength
60
+ AKT3 (HGNC:393),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
61
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
62
+ AKT3 (HGNC:393),PM2,Supporting,Absent/rare from controls in an ethnically-matched cohort population sample ( ≥1).,Disease-specific
63
+ AKT3 (HGNC:393),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
64
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
65
+ AKT3 (HGNC:393),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,NA
66
+ AKT3 (HGNC:393),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
67
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
68
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
69
+ AKT3 (HGNC:393),PM5,Moderate,No change.,None
70
+ AKT3 (HGNC:393),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
71
+ AKT3 (HGNC:393),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
72
+ Note: May be used as stronger evidence with increasing segregation data.",NA
73
+ AKT3 (HGNC:393),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,
74
+ AKT3 (HGNC:393),PP2,Supporting,"Missense constraint computed in ExAC/gnomAD was utilized. Award PP2 if the z-score > 3.09. (applicable to MTOR, PIK3CA and AKT3 but not PIK3R2).",Disease-specific
75
+ AKT3 (HGNC:393),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
76
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",NA
77
+ AKT3 (HGNC:393),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
78
+ AKT3 (HGNC:393),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
79
+ AKT3 (HGNC:393),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
80
+ AKT3 (HGNC:393),BA1,Stand Alone,"Allele frequency (>0.0926%). An allele frequency (>0.0926%) was approved.
81
+ Note: this was adjusted from ACMG Guidelines due to maintaining the 5x threshold for benign (consistent with previously established guidelines)",Disease-specific
82
+ AKT3 (HGNC:393),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
83
+ AKT3 (HGNC:393),BS1,Strong,Allele frequency (>0.0185%). An allele frequency (>0.0185%) was approved. (Supplemental Table 3).,Disease-specific
84
+ AKT3 (HGNC:393),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
85
+ AKT3 (HGNC:393),BS2,Strong,"Award BS2 if ≥3 homozygotes present in gnomAD or ≥3 heterozygous in well phenotyped family members.
86
+ Clinical laboratories are encouraged to accumulate more than 2 (≥3) instances of well phenotyped family members before applying this strong criterion. To be considered for this point, the variant should be either germline (most common), or somatic in a relevant tissue. Homozygous occurrences in gnomAD or ExAC can also be counted for this point (≥3).",Disease-specific
87
+ AKT3 (HGNC:393),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
88
+ AKT3 (HGNC:393),BS3,Strong,Follow recommendations set forth by the SVI in conjunction with specifications added by the Brain Malformation Group for quality metrics and minimum validation controls required.,Disease-specific
89
+ AKT3 (HGNC:393),BS3,Supporting,Follow recommendations set forth by the SVI in conjunction with specifications added by the Brain Malformation Group for quality metrics and minimum validation controls required.,Strength
90
+ AKT3 (HGNC:393),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
91
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
92
+ AKT3 (HGNC:393),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
93
+ AKT3 (HGNC:393),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
94
+ AKT3 (HGNC:393),BP2,Supporting,Observed in cis or trans with a known pathogenic variant in the same gene.,None
95
+ AKT3 (HGNC:393),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
96
+ AKT3 (HGNC:393),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
97
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
98
+ AKT3 (HGNC:393),BP4,Supporting,"Award BP4 for a synonymous, intronic positions (except canonical splice sites) or non-coding variants in the UTRs, if two out of three of the splicing prediction tools predicted no impact on splicing function.
99
+ Not applicable for any variant type except for synonymous, intronic positions (except canonical splice sites) and non-coding variants in the UTRs,. This criterion can be applied when two of three splicing prediction tools predict no splicing change. The splicing prediction tools used are: varSEAK, spliceAI and MaxEntScan.",Disease-specific
100
+ AKT3 (HGNC:393),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
101
+ AKT3 (HGNC:393),BP5,Supporting,No change.,None
102
+ AKT3 (HGNC:393),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
103
+ AKT3 (HGNC:393),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
104
+ AKT3 (HGNC:393),BP7,Supporting,"For synonymous, intronic positions (except canonical splice sites) and non-coding variants in the UTRs, if the nucleotide is non-conserved award this point (PhyloP score <0.1).",Disease-specific
VCI/parsing_csr_criteria/version_csv_individual/ClinGenBrainMalformationsExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv ADDED
@@ -0,0 +1,81 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ AKT3 (HGNC:393),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",NA
8
+ AKT3 (HGNC:393),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
9
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
10
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
11
+ AKT3 (HGNC:393),PS1,Strong,No change.,None
12
+ AKT3 (HGNC:393),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
13
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
14
+ AKT3 (HGNC:393),PS2,Strong,"Award the PS2_Strong point if Criteria 1 AND Criteria 2 are fulfilled.
15
+ Criteria 1: if the variant is present at a detectable allelic fraction in a proband with the disease but is absent from parental samples with confirmed maternity and paternity.
16
+ Criteria 2: can also be awarded if the variant is present at a detectable allele fraction in an affected tissue sample but is absent from or detected at a lower allelic fraction in another tissue (e.g. if present in 5% of brain tissue but absent from the blood or skin this point can be awarded).","Disease-specific,Strength"
17
+ AKT3 (HGNC:393),PS2,Moderate,"Award the PS2_Moderate point if Criteria 1 is fulfilled, OR if parents are not available but Criteria 2 is fulfilled.
18
+ Criteria 1: if the variant is present at a detectable allelic fraction in a proband with the disease but is absent from parental samples with confirmed maternity and paternity.
19
+ Criteria 2: can also be awarded if the variant is present at a detectable allele fraction in an affected tissue sample but is absent from or detected at a lower allelic fraction in another tissue (e.g. if present in 5% of brain tissue but absent from the blood or skin this point can be awarded).","Disease-specific,Strength"
20
+ AKT3 (HGNC:393),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
21
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
22
+ AKT3 (HGNC:393),PS3,Strong,"Follow recommendations set forth by the SVI in conjunction with specifications added by the Brain Malformation Group for quality metrics and minimum validation controls required.
23
+ Animal models are considered in a different manner. Award PS4_Strong if the animal model generated with the variant of interest expressed in neural progenitors show a complementary brain phenotype.",Disease-specific
24
+ AKT3 (HGNC:393),PS3,Moderate,"Follow recommendations set forth by the SVI in conjunction with specifications added by the Brain Malformation Group for quality metrics and minimum validation controls required.
25
+ Animal models are considered in a different manner. Award PS4_Moderate if the animal model generated with the variant of interest expressed in non-neural tissues show an increased cancer burden.",Strength
26
+ AKT3 (HGNC:393),PS3,Supporting,Follow recommendations set forth by the SVI in conjunction with specifications added by the Brain Malformation Group for quality metrics and minimum validation controls required.,Strength
27
+ AKT3 (HGNC:393),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
28
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
29
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
30
+ AKT3 (HGNC:393),PS4,Very Strong,"Points are assigned for phenotype according to (Table 2A). Phenotype criteria can only be used if the variant is absent from controls (PM2). Strength of evidence is determined by points according to (Table 2B).
31
+ PS4_VeryStrong = >16 points.","Disease-specific,Strength"
32
+ AKT3 (HGNC:393),PS4,Strong,"Points are assigned for phenotype according to (Table 2A). Phenotype criteria can only be used if the variant is absent from controls (PM2). Strength of evidence is determined by points according to (Table 2B).
33
+ PS4 = 3.5-15.99 points.",Disease-specific
34
+ AKT3 (HGNC:393),PS4,Moderate,"Points are assigned for phenotype according to (Table 2A). Phenotype criteria can only be used if the variant is absent from controls (PM2). Strength of evidence is determined by points according to (Table 2B).
35
+ PS4_Moderate = 1.5-3.49 points.",Strength
36
+ AKT3 (HGNC:393),PS4,Supporting,"The prevalence of the variant in affected individuals is significantly increased compared with the prevalence in controls. Points are assigned for phenotype according to (Supplementary Table 3). Strength of evidence is determined by points according to (Supplementary Table 2).
37
+ PS4_Supporting = 0.5 - 1.49 points.",Strength
38
+ AKT3 (HGNC:393),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
39
+ AKT3 (HGNC:393),PM1,Supporting,Residues affecting critical functional domains provided in Table 4 for each gene.,Strength
40
+ AKT3 (HGNC:393),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
41
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
42
+ AKT3 (HGNC:393),PM2,Supporting,Absent/rare from controls in an ethnically-matched cohort population sample ( ≥1).,Disease-specific
43
+ AKT3 (HGNC:393),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
44
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
45
+ AKT3 (HGNC:393),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,NA
46
+ AKT3 (HGNC:393),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
47
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
48
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
49
+ AKT3 (HGNC:393),PM5,Moderate,No change.,None
50
+ AKT3 (HGNC:393),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
51
+ AKT3 (HGNC:393),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
52
+ Note: May be used as stronger evidence with increasing segregation data.",NA
53
+ AKT3 (HGNC:393),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,
54
+ AKT3 (HGNC:393),PP2,Supporting,"Missense constraint computed in ExAC/gnomAD was utilized. Award PP2 if the z-score > 3.09. (applicable to MTOR, PIK3CA and AKT3 but not PIK3R2).",Disease-specific
55
+ AKT3 (HGNC:393),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
56
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",NA
57
+ AKT3 (HGNC:393),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
58
+ AKT3 (HGNC:393),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
59
+ AKT3 (HGNC:393),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
60
+ AKT3 (HGNC:393),BA1,Stand Alone,Allele frequency (≥0.185%).,Disease-specific
61
+ AKT3 (HGNC:393),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
62
+ AKT3 (HGNC:393),BS1,Strong,Allele frequency (≥0.037%).,Disease-specific
63
+ AKT3 (HGNC:393),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
64
+ AKT3 (HGNC:393),BS2,Strong,Award BS2 if ≥3 homozygotes present in gnomAD or well phenotyped family members.,Disease-specific
65
+ AKT3 (HGNC:393),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
66
+ AKT3 (HGNC:393),BS3,Strong,Follow recommendations set forth by the SVI in conjunction with specifications added by the Brain Malformation Group for quality metrics and minimum validation controls required.,Disease-specific
67
+ AKT3 (HGNC:393),BS3,Supporting,Follow recommendations set forth by the SVI in conjunction with specifications added by the Brain Malformation Group for quality metrics and minimum validation controls required.,Strength
68
+ AKT3 (HGNC:393),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
69
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
70
+ AKT3 (HGNC:393),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
71
+ AKT3 (HGNC:393),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
72
+ AKT3 (HGNC:393),BP2,Supporting,No change.,None
73
+ AKT3 (HGNC:393),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
74
+ AKT3 (HGNC:393),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
75
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
76
+ AKT3 (HGNC:393),BP4,Supporting,"Award BP4 for a synonymous, intronic positions (except canonical splice sites) or non-coding variants in the UTRs, if two out of three of the splicing prediction tools predicted no impact on splicing function.",Disease-specific
77
+ AKT3 (HGNC:393),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
78
+ AKT3 (HGNC:393),BP5,Supporting,No change.,None
79
+ AKT3 (HGNC:393),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
80
+ AKT3 (HGNC:393),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
81
+ AKT3 (HGNC:393),BP7,Supporting,"For synonymous, intronic positions (except canonical splice sites) and non-coding variants in the UTRs, if the nucleotide is non-conserved award this point (PhyloP score <0.1).",Disease-specific
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion2_version=2.0.0.csv ADDED
@@ -0,0 +1,108 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ CDH1 (HGNC:1748),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7)
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact
7
+  • Use caution in the presence of multiple transcripts",
8
+ CDH1 (HGNC:1748),PVS1,Very Strong,"Per ClinGen SVI guidelines with the exception of canonical splice sites
9
+
10
+
11
+
12
+
13
+ Apply to initiation codon variant",
14
+ CDH1 (HGNC:1748),PVS1,Strong,"Per ClinGen SVI guidelines
15
+ Other CDH1 caveats:
16
+
17
+
18
+
19
+
20
+ Use the strong strength of evidence for canonical splice sites
21
+
22
+
23
+ CDH1 Exonic deletions or tandem duplications of in-frame exons
24
+
25
+
26
+ Truncations in NMD-resistant zone located upstream the most 3’ well-characterized pathogenic variant c.2506G>T (p.Glu836*). Use PVS1_moderate if premature stop is downstream of this variant",
27
+ CDH1 (HGNC:1748),PVS1,Moderate,"Per ClinGen SVI guidelines
28
+ Other CDH1 caveats:
29
+
30
+
31
+
32
+
33
+ G to non-G variants disrupting the last nucleotide of an exon
34
+
35
+
36
+ Canonical splice sites located in exons demonstrated experimentally to result in in-frame partial skipping/insertion (e.g., Exon 3 donor site)",
37
+ CDH1 (HGNC:1748),PVS1,Supporting,Per ClinGen SVI guidelines,
38
+ CDH1 (HGNC:1748),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
39
+ Example: Val->Leu caused by either G>C or G>T in the same codon
40
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level",
41
+ CDH1 (HGNC:1748),PS1,Strong,Per original ACMG/AMP guidelines,
42
+ CDH1 (HGNC:1748),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history
43
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity",
44
+ CDH1 (HGNC:1748),PS2,Very Strong,≥Two patients with DGC &/or LBC w/ parental confirmation,
45
+ CDH1 (HGNC:1748),PS2,Strong,One patient with DGC &/or LBC w/ parental confirmation,
46
+ CDH1 (HGNC:1748),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product
47
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established",
48
+ CDH1 (HGNC:1748),PS3,Strong,RNA assay demonstrating abnormal out-of-frame transcripts,
49
+ CDH1 (HGNC:1748),PS3,Supporting,RNA assay demonstrating abnormal in-frame transcripts,
50
+ CDH1 (HGNC:1748),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls
51
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
52
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
53
+ CDH1 (HGNC:1748),PS4,Very Strong,Sixteen families meet HDGC criteria,
54
+ CDH1 (HGNC:1748),PS4,Strong,Four families meet HDGC criteria,
55
+ CDH1 (HGNC:1748),PS4,Moderate,Two families meet HDGC criteria,
56
+ CDH1 (HGNC:1748),PS4,Supporting,One family meets HDGC criteria,
57
+ CDH1 (HGNC:1748),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation,NA
58
+ CDH1 (HGNC:1748),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium
59
+ Caveat: Population data for indels may be poorly called by next generation sequencing",
60
+ CDH1 (HGNC:1748),PM2,Moderate,"Less than one out of 100,000 alleles in gnomAD cohort; if present in >=2 individuals, must be present in less than one out of 50,000 alleles within a sub-population",
61
+ CDH1 (HGNC:1748),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
62
+ Note: This requires testing of parents (or offspring) to determine phase",NA
63
+ CDH1 (HGNC:1748),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
64
+ CDH1 (HGNC:1748),PM4,Moderate,Per original ACMG/AMP guidelines,
65
+ CDH1 (HGNC:1748),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before
66
+ Example: Arg156His is pathogenic; now you observe Arg156Cys
67
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level",NA
68
+ CDH1 (HGNC:1748),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity",
69
+ CDH1 (HGNC:1748),PM6,Very Strong,≥Four patients with DGC &/or LBC w/o parental confirmation,
70
+ CDH1 (HGNC:1748),PM6,Strong,≥Two patients with DGC &/or LBC w/o parental confirmation,
71
+ CDH1 (HGNC:1748),PM6,Moderate,One patient with DGC &/or LBC w/o parental confirmation,
72
+ CDH1 (HGNC:1748),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease
73
+ Note: May be used as stronger evidence with increasing segregation data",
74
+ CDH1 (HGNC:1748),PP1,Strong,≥Seven meioses across ≥2 families,
75
+ CDH1 (HGNC:1748),PP1,Moderate,Five-six meioses across ≥1 families,
76
+ CDH1 (HGNC:1748),PP1,Supporting,Three-four meioses across ≥1 families,
77
+ CDH1 (HGNC:1748),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease,NA
78
+ CDH1 (HGNC:1748),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc)
79
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
80
+ CDH1 (HGNC:1748),PP3,Moderate,Variants affecting the same splice site as a well-characterized variant with similar or worse in silico/ RNA predictions,
81
+ CDH1 (HGNC:1748),PP3,Supporting,"At least three in silico splicing predictors in agreement (.Human Splicing Finder (HSF), Maximum Entropy (MaxEnt), Berkeley Drosophilia Genome Project (BDGP), or ESEfinder)",
82
+ CDH1 (HGNC:1748),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology,NA
83
+ CDH1 (HGNC:1748),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
84
+ CDH1 (HGNC:1748),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium",
85
+ CDH1 (HGNC:1748),BA1,Stand Alone,MAF cutoff of 0.2%,
86
+ CDH1 (HGNC:1748),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
87
+ CDH1 (HGNC:1748),BS1,Stand Alone,MAF cutoff of 0.1%,
88
+ CDH1 (HGNC:1748),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder with full penetrance expected at an early age",
89
+ CDH1 (HGNC:1748),BS2,Strong,"Variant seen in ≥10 individuals w/o DCG, SRC tumors, or LBC & whose families do not suggest HDGC",
90
+ CDH1 (HGNC:1748),BS2,Supporting,"Variant seen in ≥3 individuals w/o DCG, SRC tumors, or LBC & whose families do not suggest HDGC",
91
+ CDH1 (HGNC:1748),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
92
+ CDH1 (HGNC:1748),BS3,Strong,Functional RNA studies demonstrating no impact on transcript composition,
93
+ CDH1 (HGNC:1748),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family
94
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation",
95
+ CDH1 (HGNC:1748),BS4,Strong,Per original ACMG/AMP guidelines,
96
+ CDH1 (HGNC:1748),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease,NA
97
+ CDH1 (HGNC:1748),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
98
+ CDH1 (HGNC:1748),BP2,Strong,"Variant observed in trans w/known pathogenic variant (phase confirmed) OR observed in the homozygous state in individual w/o personal &/or family history of DGC, LBC, or SRC tumors",
99
+ CDH1 (HGNC:1748),BP2,Supporting,Variant is observed in cis (or phase is unknown) w/ a pathogenic variant,
100
+ CDH1 (HGNC:1748),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
101
+ CDH1 (HGNC:1748),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
102
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
103
+ CDH1 (HGNC:1748),BP4,Supporting,"Splicing predictions only. At least three in silico splicing predictors in agreement (Human Splicing Finder (HSF), Maximum Entropy (MaxEnt), Berkeley Drosophilia Genome Project (BDGP), or ESEfinder)",
104
+ CDH1 (HGNC:1748),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
105
+ CDH1 (HGNC:1748),BP5,Supporting,Per original ACMG/AMP guidelines,
106
+ CDH1 (HGNC:1748),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
107
+ CDH1 (HGNC:1748),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
108
+ CDH1 (HGNC:1748),BP7,Supporting,Synonymous variants where nucleotide is not highly conserved; variant is the reference nucleotide in one primate and/or >3 mammal species,
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion3.1_version=3.1.0.csv ADDED
@@ -0,0 +1,101 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ CDH1 (HGNC:1748),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ CDH1 (HGNC:1748),PVS1,Very Strong,Per modified CDH1 PVS1 decision tree.,
9
+ CDH1 (HGNC:1748),PVS1,Strong,"Per modified CDH1 PVS1 decision tree.
10
+ Other CDH1 caveats:
11
+
12
+
13
+
14
+
15
+ Use PVS1_Strong as the default strength of evidence for canonical splice site variants and follow the site-specific recommendations in the splicing table.
16
+
17
+
18
+ CDH1 Exonic deletions or tandem duplications of in-frame exons (exon 4,5,8,9,12,13,15).",
19
+ CDH1 (HGNC:1748),PVS1,Moderate,"Per modified CDH1 PVS1 decision tree.
20
+ Other CDH1 caveats:
21
+
22
+
23
+
24
+
25
+ G to non-G variants disrupting the last nucleotide of an exon.
26
+
27
+
28
+ Canonical splice sites predicted or demonstrated experimentally to result in in-frame partial skipping/insertion (e.g., Exon 3 donor site).",
29
+ CDH1 (HGNC:1748),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
30
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
31
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",NA
32
+ CDH1 (HGNC:1748),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
33
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
34
+ CDH1 (HGNC:1748),PS2,Very Strong,≥Two patients meet the HDGC individual phenotype criteria w/ parental confirmation.,
35
+ CDH1 (HGNC:1748),PS2,Strong,One patient meets the HDGC individual phenotype criteria w/ parental confirmation.,
36
+ CDH1 (HGNC:1748),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
37
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
38
+ CDH1 (HGNC:1748),PS3,Strong,RNA assay demonstrating abnormal out-of-frame transcripts.,
39
+ CDH1 (HGNC:1748),PS3,Moderate,RNA assay demonstrating abnormal in-frame transcript.,
40
+ CDH1 (HGNC:1748),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
41
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
42
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
43
+ CDH1 (HGNC:1748),PS4,Very Strong,≥Sixteen families meet HDGC criteria.,
44
+ CDH1 (HGNC:1748),PS4,Strong,Four - Fifteen families meet HDGC criteria.,
45
+ CDH1 (HGNC:1748),PS4,Moderate,Two or three families meet HDGC criteria.,
46
+ CDH1 (HGNC:1748),PS4,Supporting,One family meets HDGC criteria.,
47
+ CDH1 (HGNC:1748),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
48
+ CDH1 (HGNC:1748),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
49
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
50
+ CDH1 (HGNC:1748),PM2,Supporting,"≤ One out of 100,000 alleles in gnomAD cohort; if present in ≥2 individuals within a subpopulation, must be present in ≤ One out of 50,000 alleles.",
51
+ CDH1 (HGNC:1748),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
52
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
53
+ CDH1 (HGNC:1748),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
54
+ CDH1 (HGNC:1748),PM4,Moderate,"Only apply to stop-loss variants
55
+ Variant example: CDH1 c.2647T>C (p.Ter883Glnext*29).",
56
+ CDH1 (HGNC:1748),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
57
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
58
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
59
+ CDH1 (HGNC:1748),PM5,Supporting,PM5_supporting is applicable to nonsense and frameshift variants that are predicted/proved to undergo NMD or located upstream of the last known pathogenic truncating variant. Site-specific recommendations for the application of PM5_Supporting for canonical splicing variants.,
60
+ CDH1 (HGNC:1748),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
61
+ CDH1 (HGNC:1748),PM6,Very Strong,Four patients meet the HDGC individual phenotype criteria w/o parental confirmation.,
62
+ CDH1 (HGNC:1748),PM6,Strong,≥Two patients meet the HDGC individual phenotype criteria w/o parental confirmation.,
63
+ CDH1 (HGNC:1748),PM6,Moderate,One patient meets the HDGC individual phenotype criteria w/o parental confirmation,
64
+ CDH1 (HGNC:1748),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
65
+ Note: May be used as stronger evidence with increasing segregation data.",
66
+ CDH1 (HGNC:1748),PP1,Strong,≥Seven informative meioses across ≥2 families.,
67
+ CDH1 (HGNC:1748),PP1,Moderate,Five-six informative meioses across ≥1 family.,
68
+ CDH1 (HGNC:1748),PP1,Supporting,Three-four informative meioses across ≥1 family.,
69
+ CDH1 (HGNC:1748),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
70
+ CDH1 (HGNC:1748),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
71
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
72
+ CDH1 (HGNC:1748),PP3,Moderate,Variants affecting the same splice site as a well-characterized variant with similar or worse in silico/ RNA predictions.,
73
+ CDH1 (HGNC:1748),PP3,Supporting,"At least three in silico splicing predictors in agreement (SpliceAI, MaxEntScan, SSF, GeneSplicer, HSF, TraP, varSEAK).",
74
+ CDH1 (HGNC:1748),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
75
+ CDH1 (HGNC:1748),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
76
+ CDH1 (HGNC:1748),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
77
+ CDH1 (HGNC:1748),BA1,Stand Alone,MAF cutoff of 0.2%.,
78
+ CDH1 (HGNC:1748),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
79
+ CDH1 (HGNC:1748),BS1,Strong,MAF cutoff of 0.1%.,
80
+ CDH1 (HGNC:1748),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
81
+ CDH1 (HGNC:1748),BS2,Strong,"Variant seen in ≥10 individuals w/o GC, DGC, gSRC tumors, or LBC & whose families do not suggest HDGC.",
82
+ CDH1 (HGNC:1748),BS2,Supporting,"Variant seen in ≥3 individuals w/o GC, DGC, SRC tumors, or LBC & whose families do not suggest HDGC.",
83
+ CDH1 (HGNC:1748),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
84
+ CDH1 (HGNC:1748),BS3,Strong,Functional RNA studies demonstrating no impact on transcript composition.,
85
+ CDH1 (HGNC:1748),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
86
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
87
+ CDH1 (HGNC:1748),BS4,Strong,Per original ACMG/AMP guidelines.,
88
+ CDH1 (HGNC:1748),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
89
+ CDH1 (HGNC:1748),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
90
+ CDH1 (HGNC:1748),BP2,Strong,"Variant observed in trans w/known pathogenic variant (phase confirmed) OR observed in the homozygous state in individual w/o personal &/or family history of DGC, LBC, or SRC tumors.",
91
+ CDH1 (HGNC:1748),BP2,Supporting,"Variant is observed in cis (or phase is unknown) w/ a pathogenic variant
92
+ OR observed in the homozygous state in gnomAD.",
93
+ CDH1 (HGNC:1748),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
94
+ CDH1 (HGNC:1748),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
95
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
96
+ CDH1 (HGNC:1748),BP4,Supporting,"Splicing predictions only. At least three in silico splicing predictors in agreement (SpliceAI, MaxEntScan, SSF, GeneSplicer, HSF, TraP, varSEAK).",
97
+ CDH1 (HGNC:1748),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
98
+ CDH1 (HGNC:1748),BP5,Supporting,Per original ACMG/AMP guidelines.,
99
+ CDH1 (HGNC:1748),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
100
+ CDH1 (HGNC:1748),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
101
+ CDH1 (HGNC:1748),BP7,Supporting,Synonymous and intronic variants at or beyond +7 to -21 locations.,
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion3_version=3.0.0.csv ADDED
@@ -0,0 +1,101 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ CDH1 (HGNC:1748),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ CDH1 (HGNC:1748),PVS1,Very Strong,Per modified CDH1 PVS1 decision tree.,
9
+ CDH1 (HGNC:1748),PVS1,Strong,"Per modified CDH1 PVS1 decision tree.
10
+ Other CDH1 caveats:
11
+
12
+
13
+
14
+
15
+ Use PVS1_Strong as the default strength of evidence for canonical splice site variants and follow the site-specific recommendations in the splicing table.
16
+
17
+
18
+ CDH1 Exonic deletions or tandem duplications of in-frame exons (exon 4,5,8,9,12,13,15).",
19
+ CDH1 (HGNC:1748),PVS1,Moderate,"Per modified CDH1 PVS1 decision tree.
20
+ Other CDH1 caveats:
21
+
22
+
23
+
24
+
25
+ G to non-G variants disrupting the last nucleotide of an exon.
26
+
27
+
28
+ Canonical splice sites predicted or demonstrated experimentally to result in in-frame partial skipping/insertion (e.g., Exon 3 donor site).",
29
+ CDH1 (HGNC:1748),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
30
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
31
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",NA
32
+ CDH1 (HGNC:1748),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
33
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
34
+ CDH1 (HGNC:1748),PS2,Very Strong,≥Two patients meet the HDGC individual phenotype criteria w/ parental confirmation.,
35
+ CDH1 (HGNC:1748),PS2,Strong,One patient meets the HDGC individual phenotype criteria w/ parental confirmation.,
36
+ CDH1 (HGNC:1748),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
37
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
38
+ CDH1 (HGNC:1748),PS3,Strong,RNA assay demonstrating abnormal out-of-frame transcripts.,
39
+ CDH1 (HGNC:1748),PS3,Moderate,RNA assay demonstrating abnormal in-frame transcript.,
40
+ CDH1 (HGNC:1748),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
41
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
42
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
43
+ CDH1 (HGNC:1748),PS4,Very Strong,≥Sixteen families meet HDGC criteria.,
44
+ CDH1 (HGNC:1748),PS4,Strong,Four - Fifteen families meet HDGC criteria.,
45
+ CDH1 (HGNC:1748),PS4,Moderate,Two or three families meet HDGC criteria.,
46
+ CDH1 (HGNC:1748),PS4,Supporting,One family meets HDGC criteria.,
47
+ CDH1 (HGNC:1748),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
48
+ CDH1 (HGNC:1748),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
49
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
50
+ CDH1 (HGNC:1748),PM2,Supporting,"≤ One out of 100,000 alleles in gnomAD cohort; if present in ≥2 individuals within a subpopulation, must be present in ≤ One out of 50,000 alleles.",
51
+ CDH1 (HGNC:1748),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
52
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
53
+ CDH1 (HGNC:1748),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
54
+ CDH1 (HGNC:1748),PM4,Moderate,"Only apply to stop-loss variants
55
+ Variant example: CDH1 c.2647T>C (p.Ter883Glnext*29).",
56
+ CDH1 (HGNC:1748),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
57
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
58
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
59
+ CDH1 (HGNC:1748),PM5,Supporting,PM5_supporting is applicable to nonsense and frameshift variants that are predicted/proved to undergo NMD. Site-specific recommendations for the application of PM5_Supporting for canonical splicing variants are provided in the splicing table.,
60
+ CDH1 (HGNC:1748),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
61
+ CDH1 (HGNC:1748),PM6,Very Strong,Four patients meet the HDGC individual phenotype criteria w/o parental confirmation.,
62
+ CDH1 (HGNC:1748),PM6,Strong,≥Two patients meet the HDGC individual phenotype criteria w/o parental confirmation.,
63
+ CDH1 (HGNC:1748),PM6,Moderate,One patient meets the HDGC individual phenotype criteria w/o parental confirmation,
64
+ CDH1 (HGNC:1748),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
65
+ Note: May be used as stronger evidence with increasing segregation data.",
66
+ CDH1 (HGNC:1748),PP1,Strong,≥Seven informative meioses across ≥2 families.,
67
+ CDH1 (HGNC:1748),PP1,Moderate,Five-six informative meioses across ≥1 family.,
68
+ CDH1 (HGNC:1748),PP1,Supporting,Three-four informative meioses across ≥1 family.,
69
+ CDH1 (HGNC:1748),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
70
+ CDH1 (HGNC:1748),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
71
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
72
+ CDH1 (HGNC:1748),PP3,Moderate,Variants affecting the same splice site as a well-characterized variant with similar or worse in silico/ RNA predictions.,
73
+ CDH1 (HGNC:1748),PP3,Supporting,"At least three in silico splicing predictors in agreement (SpliceAI, MaxEntScan, SSF, GeneSplicer, HSF, TraP, varSEAK).",
74
+ CDH1 (HGNC:1748),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
75
+ CDH1 (HGNC:1748),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
76
+ CDH1 (HGNC:1748),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
77
+ CDH1 (HGNC:1748),BA1,Stand Alone,MAF cutoff of 0.2%.,
78
+ CDH1 (HGNC:1748),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
79
+ CDH1 (HGNC:1748),BS1,Strong,MAF cutoff of 0.1%.,
80
+ CDH1 (HGNC:1748),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
81
+ CDH1 (HGNC:1748),BS2,Strong,"Variant seen in ≥10 individuals w/o GC, DGC, gSRC tumors, or LBC & whose families do not suggest HDGC.",
82
+ CDH1 (HGNC:1748),BS2,Supporting,"Variant seen in ≥3 individuals w/o GC, DGC, SRC tumors, or LBC & whose families do not suggest HDGC.",
83
+ CDH1 (HGNC:1748),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
84
+ CDH1 (HGNC:1748),BS3,Strong,Functional RNA studies demonstrating no impact on transcript composition.,
85
+ CDH1 (HGNC:1748),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
86
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
87
+ CDH1 (HGNC:1748),BS4,Strong,Per original ACMG/AMP guidelines.,
88
+ CDH1 (HGNC:1748),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
89
+ CDH1 (HGNC:1748),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
90
+ CDH1 (HGNC:1748),BP2,Strong,"Variant observed in trans w/known pathogenic variant (phase confirmed) OR observed in the homozygous state in individual w/o personal &/or family history of DGC, LBC, or SRC tumors.",
91
+ CDH1 (HGNC:1748),BP2,Supporting,"Variant is observed in cis (or phase is unknown) w/ a pathogenic variant
92
+ OR observed in the homozygous state in gnomAD.",
93
+ CDH1 (HGNC:1748),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
94
+ CDH1 (HGNC:1748),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
95
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
96
+ CDH1 (HGNC:1748),BP4,Supporting,"Splicing predictions only. At least three in silico splicing predictors in agreement (SpliceAI, MaxEntScan, SSF, GeneSplicer, HSF, TraP, varSEAK).",
97
+ CDH1 (HGNC:1748),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
98
+ CDH1 (HGNC:1748),BP5,Supporting,Per original ACMG/AMP guidelines.,
99
+ CDH1 (HGNC:1748),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
100
+ CDH1 (HGNC:1748),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
101
+ CDH1 (HGNC:1748),BP7,Supporting,Synonymous and intronic variants at or beyond +7 to -21 locations.,
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCDH1ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforCDH1Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,101 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ CDH1 (HGNC:1748),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ CDH1 (HGNC:1748),PVS1,Very Strong,Per ClinGen SVI guidelines with the exception of canonical splice sites.,
9
+ CDH1 (HGNC:1748),PVS1,Strong,"Per modified CDH1 PVS1 decision tree. Other CDH1 caveats:
10
+
11
+
12
+
13
+
14
+ Use the strong strength of evidence for canonical splice sites.
15
+
16
+
17
+ CDH1 Exonic deletions or tandem duplications of in-frame exons.
18
+
19
+
20
+ Truncations in NMD-resistant zone located upstream the most 3’ well-characterized pathogenic variant c.2506G>T (p.Glu836*). Use PVS1_moderate if premature stop is downstream of this variant.",
21
+ CDH1 (HGNC:1748),PVS1,Moderate,"Per modified CDH1 PVS1 decision tree. Other CDH1 caveats:
22
+
23
+
24
+
25
+
26
+ G to non-G variants disrupting the last nucleotide of an exon.
27
+
28
+
29
+ Canonical splice sites located in exons demonstrated experimentally to result in in-frame partial skipping/insertion (e.g., Exon 3 donor site).",
30
+ CDH1 (HGNC:1748),PVS1,Supporting,Per ClinGen SVI guidelines,
31
+ CDH1 (HGNC:1748),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
32
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
33
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
34
+ CDH1 (HGNC:1748),PS1,Strong,Per original ACMG/AMP guidelines.,
35
+ CDH1 (HGNC:1748),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
36
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
37
+ CDH1 (HGNC:1748),PS2,Very Strong,≥Two patients with DGC &/or LBC w/ parental confirmation.,
38
+ CDH1 (HGNC:1748),PS2,Strong,One patient with DGC &/or LBC w/ parental confirmation.,
39
+ CDH1 (HGNC:1748),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
40
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
41
+ CDH1 (HGNC:1748),PS3,Strong,RNA assay demonstrating abnormal out-of-frame transcripts.,
42
+ CDH1 (HGNC:1748),PS3,Supporting,RNA assay demonstrating abnormal in-frame transcripts.,
43
+ CDH1 (HGNC:1748),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
44
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
45
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
46
+ CDH1 (HGNC:1748),PS4,Very Strong,Sixteen families meet HDGC criteria.,
47
+ CDH1 (HGNC:1748),PS4,Strong,Four families meet HDGC criteria.,
48
+ CDH1 (HGNC:1748),PS4,Moderate,Two families meet HDGC criteria.,
49
+ CDH1 (HGNC:1748),PS4,Supporting,One family meets HDGC criteria.,
50
+ CDH1 (HGNC:1748),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
51
+ CDH1 (HGNC:1748),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
52
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
53
+ CDH1 (HGNC:1748),PM2,Moderate,"<One out of 100,000 alleles in gnomAD cohort; if present in >2 individuals, must be present in <One out of 50,000 alleles within a sub-population.",
54
+ CDH1 (HGNC:1748),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
55
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
56
+ CDH1 (HGNC:1748),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
57
+ CDH1 (HGNC:1748),PM4,Moderate,Per original ACMG/AMP guidelines.,
58
+ CDH1 (HGNC:1748),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
59
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
60
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",NA
61
+ CDH1 (HGNC:1748),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
62
+ CDH1 (HGNC:1748),PM6,Very Strong,≥Four patients with DGC &/or LBC w/o parental confirmation.,
63
+ CDH1 (HGNC:1748),PM6,Strong,≥Two patients with DGC &/or LBC w/o parental confirmation.,
64
+ CDH1 (HGNC:1748),PM6,Moderate,One patient with DGC &/or LBC w/o parental confirmation.,
65
+ CDH1 (HGNC:1748),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
66
+ Note: May be used as stronger evidence with increasing segregation data.",
67
+ CDH1 (HGNC:1748),PP1,Strong,≥Seven meioses across ≥2 families.,
68
+ CDH1 (HGNC:1748),PP1,Moderate,Five-six informative meioses across ≥1 family.,
69
+ CDH1 (HGNC:1748),PP1,Supporting,Three-four informative meioses across ≥1 family.,
70
+ CDH1 (HGNC:1748),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
71
+ CDH1 (HGNC:1748),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
72
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
73
+ CDH1 (HGNC:1748),PP3,Moderate,Variants affecting the same splice site as a well-characterized variant with similar or worse in silico/ RNA predictions.,
74
+ CDH1 (HGNC:1748),PP3,Supporting,"At least three in silico splicing predictors in agreement (.Human Splicing Finder (HSF), Maximum Entropy (MaxEnt), Berkeley Drosophilia Genome Project (BDGP), or ESEfinder).",
75
+ CDH1 (HGNC:1748),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
76
+ CDH1 (HGNC:1748),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
77
+ CDH1 (HGNC:1748),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
78
+ CDH1 (HGNC:1748),BA1,Stand Alone,MAF cutoff of 0.2%.,
79
+ CDH1 (HGNC:1748),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
80
+ CDH1 (HGNC:1748),BS1,Strong,MAF cutoff of 0.1%.,
81
+ CDH1 (HGNC:1748),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
82
+ CDH1 (HGNC:1748),BS2,Strong,"Variant seen in ≥10 individuals w/o DCG, SRC tumors, or LBC & whose families do not suggest HDGC.",
83
+ CDH1 (HGNC:1748),BS2,Supporting,"Variant seen in ≥3 individuals w/o DCG, SRC tumors, or LBC & whose families do not suggest HDGC.",
84
+ CDH1 (HGNC:1748),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
85
+ CDH1 (HGNC:1748),BS3,Strong,Functional RNA studies demonstrating no impact on transcript composition.,
86
+ CDH1 (HGNC:1748),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
87
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
88
+ CDH1 (HGNC:1748),BS4,Strong,Per original ACMG/AMP guidelines.,
89
+ CDH1 (HGNC:1748),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
90
+ CDH1 (HGNC:1748),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
91
+ CDH1 (HGNC:1748),BP2,Strong,"Variant observed in trans w/known pathogenic variant (phase confirmed) OR observed in the homozygous state in individual w/o personal &/or family history of DGC, LBC, or SRC tumors.",
92
+ CDH1 (HGNC:1748),BP2,Supporting,Variant is observed in cis (or phase is unknown) w/ a pathogenic variant.,
93
+ CDH1 (HGNC:1748),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
94
+ CDH1 (HGNC:1748),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
95
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
96
+ CDH1 (HGNC:1748),BP4,Supporting,"Splicing predictions only. At least three in silico splicing predictors in agreement (Human Splicing Finder (HSF), Maximum Entropy (MaxEnt), Berkeley Drosophilia Genome Project (BDGP), or ESEfinder).",
97
+ CDH1 (HGNC:1748),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
98
+ CDH1 (HGNC:1748),BP5,Supporting,Per original ACMG/AMP guidelines.,
99
+ CDH1 (HGNC:1748),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
100
+ CDH1 (HGNC:1748),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
101
+ CDH1 (HGNC:1748),BP7,Supporting,Synonymous variants where nucleotide is not highly conserved; variant is the reference nucleotide in one primate and/or >3 mammal species.,
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesVersion1_version=1.0.0.csv ADDED
@@ -0,0 +1,72 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ MYH7 (HGNC:7577),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ MYH7 (HGNC:7577),PVS1,Moderate,Null Variant in gene with evidence supporting LOF as disease mechanism,Modified rule strength
9
+ MYH7 (HGNC:7577),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
10
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
11
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
12
+ MYH7 (HGNC:7577),PS1,Strong,Different nucleotide change (same amino acid) as a previously established pathogenic variant,No change
13
+ MYH7 (HGNC:7577),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
14
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
15
+ MYH7 (HGNC:7577),PS2,Strong,De novo (paternity confirmed) in a patient with disease and no family history,"Disease-specific,Gene-specific"
16
+ MYH7 (HGNC:7577),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
17
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
18
+ MYH7 (HGNC:7577),PS3,Strong,Functional studies of mammalian knock-in models supportive of a damaging effect on the gene or gene product,"Disease-specific,Gene-specific"
19
+ MYH7 (HGNC:7577),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
20
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
21
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
22
+ MYH7 (HGNC:7577),PS4,Strong,Prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls -OR- Variant identified in ≥15 probands with consistent phenotypes,"Disease-specific,Gene-specific"
23
+ MYH7 (HGNC:7577),PS4,Moderate,Variant identified in >=6 probands with consistent phenotypes,Modified rule strength
24
+ MYH7 (HGNC:7577),PS4,Supporting,Variant identified in >=2 probands with consistent phenotypes,Modified rule strength
25
+ MYH7 (HGNC:7577),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
26
+ MYH7 (HGNC:7577),PM1,Moderate,Hotspot/est. functional domain (amino acids 181-937) without benign variation,"Disease-specific,Gene-specific"
27
+ MYH7 (HGNC:7577),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
28
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
29
+ MYH7 (HGNC:7577),PM2,Moderate,Absent/extremely rare (<0.004%) from large population studies.,"Disease-specific,Gene-specific"
30
+ MYH7 (HGNC:7577),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
31
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
32
+ MYH7 (HGNC:7577),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
33
+ MYH7 (HGNC:7577),PM4,Moderate,Protein length changes due to in-frame deletions/insertions of any size in a non-repeat region or stop-loss variants,"Disease-specific,Gene-specific"
34
+ MYH7 (HGNC:7577),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
35
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
36
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
37
+ MYH7 (HGNC:7577),PM5,Moderate,Missense change at an amino acid residue where a different missense change previously established as pathogenic,No change
38
+ MYH7 (HGNC:7577),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
39
+ MYH7 (HGNC:7577),PM6,Moderate,Confirmed de novo without confirmation of paternity,"Disease-specific,Gene-specific"
40
+ MYH7 (HGNC:7577),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
41
+ Note: May be used as stronger evidence with increasing segregation data.",
42
+ MYH7 (HGNC:7577),PP1,Strong,Variant segregates with >= 7 meioses,Modified rule strength
43
+ MYH7 (HGNC:7577),PP1,Moderate,Variant segragates in >=5 meioses,Modified rule strength
44
+ MYH7 (HGNC:7577),PP1,Supporting,Variant segragates in >=3 meioses,"Disease-specific,Gene-specific"
45
+ MYH7 (HGNC:7577),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
46
+ MYH7 (HGNC:7577),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
47
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
48
+ MYH7 (HGNC:7577),PP3,Supporting,Multiple lines of computational evidence support a deleterious effect on the gene or gene product,No change
49
+ MYH7 (HGNC:7577),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
50
+ MYH7 (HGNC:7577),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
51
+ MYH7 (HGNC:7577),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
52
+ MYH7 (HGNC:7577),BA1,Stand Alone,Allele frequency is >= 0.1% based on the filtering allele frequency (FAF) in ExAC,"Disease-specific,Gene-specific"
53
+ MYH7 (HGNC:7577),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
54
+ MYH7 (HGNC:7577),BS1,Strong,Allele frequency is >=0.02% based on the filtering allele frequency (FAF) in ExAC provided there is no conflicting information,"Disease-specific,Gene-specific"
55
+ MYH7 (HGNC:7577),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",NA
56
+ MYH7 (HGNC:7577),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
57
+ MYH7 (HGNC:7577),BS3,Strong,Functional studies of mammalian knock-in models supportive of no damaging effect on protein function or splicing,No change
58
+ MYH7 (HGNC:7577),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
59
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
60
+ MYH7 (HGNC:7577),BS4,Strong,Non-segregation in affected members of a family,"Disease-specific,Gene-specific"
61
+ MYH7 (HGNC:7577),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
62
+ MYH7 (HGNC:7577),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
63
+ MYH7 (HGNC:7577),BP2,Supporting,Observed as comp het (in trans) or double het in genes with overlapping function (e.g. sarcomere genes) without increased disease severity -OR- Observed in cis with a pathogenic variant in any inheritance pattern,"Disease-specific,Gene-specific"
64
+ MYH7 (HGNC:7577),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
65
+ MYH7 (HGNC:7577),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
66
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
67
+ MYH7 (HGNC:7577),BP4,Supporting,Multiple lines of computational evidence suggest no impact on gene or gene product,No change
68
+ MYH7 (HGNC:7577),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
69
+ MYH7 (HGNC:7577),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease,"Disease-specific,Gene-specific"
70
+ MYH7 (HGNC:7577),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
71
+ MYH7 (HGNC:7577),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
72
+ MYH7 (HGNC:7577),BP7,Supporting,A silent variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site -AND- the nucleotide is not highly conserved,No change
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACTC1Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,871 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ ACTC1 (HGNC:143),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",NA
8
+ ACTC1 (HGNC:143),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
9
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
10
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
11
+ ACTC1 (HGNC:143),PS1,Strong,"No cardiomyopathy specifications. Apply as outlined by Richards
12
+ et al
13
+ . 2015
14
+ 1
15
+ .
16
+
17
+
18
+ Example of when rule should NOT be applied. NM_000256.3(
19
+ MYBPC3
20
+ ): c.2308G>A (p.Asp770Asn) has an established impact on splicing leading to nonsense mediated decay (NMD) and should not be used to provide evidence for other variants observed to result in the same amino acid change.",No change
21
+ ACTC1 (HGNC:143),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
22
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
23
+ ACTC1 (HGNC:143),PS2,Strong,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
24
+ 2
25
+ .
26
+
27
+
28
+ For most cardiomyopathies, it is recommended to default to
29
+ Phenotype consistency: “Phenotype consistent with gene but not highly specific”
30
+ . Clinical judgment is required for shifting to a higher or lower category. 
31
+
32
+
33
+ For use as a STRONG or VERY STRONG criterion, ideally parents have been thoroughly clinically evaluated without evidence of cardiomyopathy (ideally using a combination of ECG and echocardiogram or cardiac MRI for maximum sensitivity).
34
+
35
+
36
+ A family history consistent with
37
+ de novo
38
+ inheritance should not have any clinical signs or symptoms suggestive of cardiomyopathy in a 1
39
+ st
40
+ or 2
41
+ nd
42
+ degree relative, for example: 
43
+
44
+
45
+
46
+
47
+ Sudden death under 60 years of age
48
+
49
+
50
+ Heart transplant
51
+
52
+
53
+ Implantable cardiac defibrillator (ICD) under 60 years of age
54
+
55
+
56
+ Features of cardiomyopathy (e.g., systolic dysfunction, hypertrophy, left ventricular enlargement in an individual without risk factors).
57
+
58
+
59
+ Other related/overlapping cardiomyopathies
60
+
61
+
62
+
63
+
64
+ Examples of non-suspicious family history may include non-specific clinical features (e.g., palpitations, syncope, borderline/inconclusive echocardiogram findings, heart attack if age appropriate and suspected to result from coronary artery disease), but every attempt should be made to clarify features. 
65
+
66
+
67
+ Generally, this criterion is only applicable in the ABSENCE of any other possible disease-causing variants.  If other pathogenic or likely pathogenic variants are present, consider decreasing points assigned or overall weight.",Disease-specific
68
+ ACTC1 (HGNC:143),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
69
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
70
+ ACTC1 (HGNC:143),PS3,Strong,"In vitro splicing assays (e.g., RNA studies)
71
+
72
+
73
+ In vitro
74
+ splicing assays may be considered as 
75
+ STRONG
76
+ evidence, providing the following criteria are met.
77
+
78
+
79
+
80
+
81
+ Prior knowledge of predominant transcripts in cardiac tissue
82
+
83
+
84
+
85
+
86
+ Analysis undertaken using RNA extracted from cardiac tissue from the individual with the variant
87
+
88
+
89
+ Analysis undertaken using RNA extracted from whole blood providing the relevant transcripts (isoforms) are expressed in blood and are at sufficient levels to assess splice disruption.
90
+
91
+
92
+ Assay shows a clear, reproducible and convincing effect on splicing (i.e. a distinct splice product, present at a level comparable to the splice product from the wild-type allele), which is not observed in controls
93
+
94
+
95
+
96
+
97
+ Confirmation of abnormal splice product by Sanger sequencing
98
+
99
+
100
+
101
+
102
+ NOTE:
103
+ Mini-gene assay in non-patient derived cell lines are NOT considered to provide STRONG evidence.
104
+
105
+
106
+ NOTE:
107
+   Whether to activate this rule needs to be reconciled with the variant spectrum and disease mechanism for the gene at hand (i.e., consider whether the effect is likely to lead to LOF or an in-frame alteration and whether this type of effect is expected to be disease causing) (Abou Tayoun
108
+ et al.
109
+ 2018
110
+ 3
111
+ ).",Disease-specific
112
+ ACTC1 (HGNC:143),PS3,Moderate,"In vivo models (e.g., variant knock-in animal models)
113
+
114
+
115
+ Mammalian variant-specific knock-in animal models that produce a phenotype consistent with the clinical phenotype in humans (e.g., structural and/or functional cardiac abnormalities, premature death, arrhythmia) may be considered as
116
+ MODERATE
117
+ evidence
118
+
119
+
120
+ NOTE:
121
+ The following assays/models do NOT meet criteria
122
+
123
+
124
+
125
+
126
+ Assays that are known to be associated with non-specific cardiac phenotypes (e.g., morpholino-induced pericardial edema in zebrafish)
127
+
128
+
129
+ In vivo evidence that is not variant specific, such as whole gene alterations (i.e., cDNA or whole gene transgenic mice and whole or partial gene knock-out mice)",Disease-specific
130
+ ACTC1 (HGNC:143),PS3,Supporting,"In vitro
131
+
132
+ assays (e.g., biochemical assays of myofilament function, motility assays, human iPSC-CM)
133
+
134
+
135
+ While some
136
+ in vitro
137
+ assays may provide evidence that a variant in a cardiomyopathy gene has an effect on protein and/or myofilament function, at present, there are no validated “gold-standard” assays that are considered to reliably predict the clinical phenotype.
138
+
139
+
140
+ As such, in the cardiomyopathy genes listed in these guidelines, data from individual
141
+ in vitro
142
+ studies are unlikely to meet the criteria required to assign this rule at more than SUPPORTING level.",Disease-specific
143
+ ACTC1 (HGNC:143),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
144
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
145
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
146
+ ACTC1 (HGNC:143),PS4,Strong,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
147
+
148
+
149
+ Cohorts used in these analyses should meet the following criteria: 
150
+
151
+
152
+
153
+
154
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
155
+
156
+
157
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
158
+
159
+
160
+
161
+
162
+
163
+
164
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
165
+
166
+
167
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
168
+
169
+
170
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
171
+
172
+
173
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
174
+
175
+
176
+ The population diversity of the case and control cohorts are broadly similar.
177
+
178
+
179
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
180
+ et al.
181
+ 2017
182
+ 5
183
+ ,
184
+ DECIPHER
185
+ ).
186
+
187
+
188
+
189
+
190
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
191
+
192
+
193
+
194
+
195
+ STRONG
196
+ evidence requires the lower bound of the 95% confidence interval (CI) around the odds ratio (OR) estimate to be
197
+ ≥20
198
+
199
+
200
+
201
+
202
+ A PS4 calculator is available at
203
+ www.cardiodb.org
204
+ .
205
+
206
+
207
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
208
+
209
+
210
+ *RELEVANT PHENOTYPES:
211
+
212
+
213
+
214
+
215
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
216
+
217
+
218
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
219
+
220
+
221
+ Additional considerations for LVNC and end-stage HCM: 
222
+
223
+
224
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
225
+ et al.
226
+ 2017
227
+ 6
228
+ ; Oechslin
229
+ et al.
230
+ 2017
231
+ 7
232
+ ; Hershberger
233
+ et al.
234
+ 2017
235
+ 8
236
+ ; Ross
237
+ et al.
238
+ 2020
239
+ 9
240
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
241
+
242
+
243
+
244
+
245
+
246
+
247
+
248
+
249
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
250
+ ACTC1 (HGNC:143),PS4,Moderate,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
251
+
252
+
253
+ Cohorts used in these analyses should meet the following criteria: 
254
+
255
+
256
+
257
+
258
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
259
+
260
+
261
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
262
+
263
+
264
+
265
+
266
+
267
+
268
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
269
+
270
+
271
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
272
+
273
+
274
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
275
+
276
+
277
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
278
+
279
+
280
+ The population diversity of the case and control cohorts are broadly similar.
281
+
282
+
283
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
284
+ et al.
285
+ 2017
286
+ 5
287
+ ,
288
+ DECIPHER
289
+ ).
290
+
291
+
292
+
293
+
294
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
295
+
296
+
297
+
298
+
299
+ MODERATE
300
+ evidence requires the lower bound of the 95% CI around the OR to be
301
+ ≥10
302
+
303
+
304
+
305
+
306
+ A PS4 calculator is available at
307
+ www.cardiodb.org
308
+ .
309
+
310
+
311
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
312
+
313
+
314
+ *RELEVANT PHENOTYPES:
315
+
316
+
317
+
318
+
319
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
320
+
321
+
322
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
323
+
324
+
325
+ Additional considerations for LVNC and end-stage HCM: 
326
+
327
+
328
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
329
+ et al.
330
+ 2017
331
+ 6
332
+ ; Oechslin
333
+ et al.
334
+ 2017
335
+ 7
336
+ ; Hershberger
337
+ et al.
338
+ 2017
339
+ 8
340
+ ; Ross
341
+ et al.
342
+ 2020
343
+ 9
344
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
345
+
346
+
347
+
348
+
349
+
350
+
351
+
352
+
353
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
354
+ ACTC1 (HGNC:143),PS4,Supporting,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
355
+
356
+
357
+ Cohorts used in these analyses should meet the following criteria: 
358
+
359
+
360
+
361
+
362
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
363
+
364
+
365
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
366
+
367
+
368
+
369
+
370
+
371
+
372
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
373
+
374
+
375
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
376
+
377
+
378
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
379
+
380
+
381
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
382
+
383
+
384
+ The population diversity of the case and control cohorts are broadly similar.
385
+
386
+
387
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
388
+ et al.
389
+ 2017
390
+ 5
391
+ ,
392
+ DECIPHER
393
+ ).
394
+
395
+
396
+
397
+
398
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
399
+
400
+
401
+
402
+
403
+ SUPPORTING
404
+ evidence requires the lower bound of the 95% CI around the OR to be
405
+ ≥5
406
+
407
+
408
+
409
+
410
+ A PS4 calculator is available at
411
+ www.cardiodb.org
412
+ .
413
+
414
+
415
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
416
+
417
+
418
+ *RELEVANT PHENOTYPES:
419
+
420
+
421
+
422
+
423
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
424
+
425
+
426
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
427
+
428
+
429
+ Additional considerations for LVNC and end-stage HCM: 
430
+
431
+
432
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
433
+ et al.
434
+ 2017
435
+ 6
436
+ ; Oechslin
437
+ et al.
438
+ 2017
439
+ 7
440
+ ; Hershberger
441
+ et al.
442
+ 2017
443
+ 8
444
+ ; Ross
445
+ et al.
446
+ 2020
447
+ 9
448
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
449
+
450
+
451
+
452
+
453
+
454
+
455
+
456
+
457
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
458
+ ACTC1 (HGNC:143),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
459
+ ACTC1 (HGNC:143),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
460
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
461
+ ACTC1 (HGNC:143),PM2,Supporting,"The values used to calculate the PM2 thresholds were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/500 or lower), where the most frequent pathogenic variant accounts for no more than 2% of cases (e.g., has an allele frequency of ≤0.02 in cases based on the upper bound of 95% CI), and where the penetrance of a pathogenic variant is expected to be at least 50% (Kelly
462
+ et al.
463
+ 2018
464
+ 10
465
+ ).
466
+
467
+
468
+ A threshold of
469
+ ≤0.00004
470
+ in the subpopulation with the highest frequency when using the upper bound of the 95% CI activates this rule.
471
+
472
+
473
+
474
+
475
+ Alternatively, this is equivalent to the variant NOT being observed more than once (≤1 allele) in gnomAD v.2.1.1 in one of the non-founder populations (e.g., absence required from the Other and Ashkenazi Jewish subpopulations).
476
+
477
+
478
+ Applying a threshold of ≤0.00004 (upper bound of 95% CI of the allele frequency in gnomAD) is equivalent to the variant being seen in a single subpopulation and that subpopulation meets any of the following:
479
+
480
+
481
+ Allele Count (AC) in Allele Number (AN)
482
+
483
+
484
+ ≤1 in ≥120,000
485
+
486
+
487
+ ≤2 in ≥160,000
488
+
489
+
490
+ ≤3 in ≥195,000
491
+
492
+
493
+ ≤4 in ≥230,000
494
+
495
+
496
+
497
+
498
+
499
+
500
+
501
+
502
+ gnomAD is the preferred database for this calculation, but currently only displays the filtering allele frequency (FAF), which is equivalent to a lower bound estimate of the 95% CI, when the upper bound is what is needed.
503
+
504
+
505
+
506
+
507
+ Confidence interval tools, such as
508
+ Confit-de-MAF
509
+ , can be used to determine the upper bound of the 95% CI of the observed allele frequency.
510
+
511
+
512
+
513
+
514
+ Due to current technical limitations of next generation sequencing technologies, minor allele frequencies for complex variants (e.g., large indels) may not be accurately represented in population databases.
515
+
516
+
517
+ Caution should be used when a variant is only identified, or over-represented, in one of the smaller gnomAD populations, as the gnomAD allele frequencies may not accurately represent the true population frequency.
518
+
519
+
520
+ Population databases may contain affected or pre-symptomatic individuals for diseases with reduced penetrance/variable onset.",Disease-specific
521
+ ACTC1 (HGNC:143),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
522
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
523
+ ACTC1 (HGNC:143),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
524
+ ACTC1 (HGNC:143),PM4,Moderate,"Strength of rule should be carefully considered and may require downgrading to SUPPORTING based on the predicted impact of the variant, including the size of the deletion/insertion, its location, and conservation of the region. 
525
+
526
+
527
+ For genes where PVS1 is not applicable (i.e., where there is no evidence that pLOF variants cause disease), consider using this rule at MODERATE or SUPPORTING strength for truncating variants that do NOT undergo nonsense mediated decay (NMD).",General recommendation
528
+ ACTC1 (HGNC:143),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
529
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
530
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
531
+ ACTC1 (HGNC:143),PM5,Moderate,"This criterion can be used at MODERATE if a different missense variant at the same codon has been classified as
532
+ pathogenic
533
+ using these modified guidelines without application of PM5.
534
+
535
+
536
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to SUPPORTING if the predicted impact is not expected to be equivalent or more severe.
537
+
538
+
539
+ PM5 should not be combined with PM1.  If both are applicable at MODERATE weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
540
+ ACTC1 (HGNC:143),PM5,Supporting,"This criterion can be considered at SUPPORTING if a different missense variant at the same codon has been classified as
541
+ likely pathogenic
542
+ using these modified guidelines without application of PM5.
543
+
544
+
545
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as likely pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to NOT APPLICABLE if the predicted impact is not expected to be equivalent or more severe.
546
+
547
+
548
+ PM5 should not be combined with PM1.  The one with the higher strength should be applied, but if both are applicable at SUPPORTING weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
549
+ ACTC1 (HGNC:143),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
550
+ ACTC1 (HGNC:143),PM6,Moderate,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
551
+ 2
552
+ .
553
+
554
+
555
+ For most cardiomyopathies, it is recommended to default to “phenotype consistent with gene but not highly specific”. Clinical judgment is required for shifting to a higher or lower phenotypic consistency. 
556
+
557
+
558
+ See PS2 for additional considerations.",Disease-specific
559
+ ACTC1 (HGNC:143),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
560
+ Note: May be used as stronger evidence with increasing segregation data.",
561
+ ACTC1 (HGNC:143),PP1,Strong,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
562
+ ≥7
563
+
564
+ segregations
565
+ (LOD score of 2.1) for
566
+ STRONG
567
+ .
568
+
569
+
570
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
571
+ 11
572
+ can be considered: 
573
+
574
+
575
+
576
+
577
+ STRONG evidence requires ≥5 segregations (LOD score of 1.5)
578
+
579
+
580
+
581
+
582
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
583
+
584
+
585
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
586
+
587
+
588
+ Important considerations include:
589
+
590
+
591
+
592
+
593
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
594
+
595
+
596
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1.
597
+
598
+
599
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
600
+
601
+
602
+ Caution is needed when distantly related (≥3
603
+ rd
604
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
605
+ ACTC1 (HGNC:143),PP1,Moderate,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
606
+ ≥5
607
+
608
+ segregations
609
+ (LOD score of 1.5) for
610
+ MODERATE
611
+ .
612
+
613
+
614
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
615
+ 11
616
+ can be considered: 
617
+
618
+
619
+
620
+
621
+ MODERATE evidence requires ≥4 segregations (LOD score of 1.2)
622
+
623
+
624
+
625
+
626
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
627
+
628
+
629
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
630
+
631
+
632
+ Important considerations include:
633
+
634
+
635
+
636
+
637
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
638
+
639
+
640
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
641
+
642
+
643
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
644
+
645
+
646
+ Caution is needed when distantly related (≥3
647
+ rd
648
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
649
+ ACTC1 (HGNC:143),PP1,Supporting,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
650
+ ≥3
651
+
652
+ segregations
653
+ (LOD score of 0.9) for
654
+ SUPPORTING
655
+ . The thresholds as proposed by Jarvik and Browning (2016)
656
+ 11
657
+ are the same at ≥3 segregations (LOD score of 0.9) for supporting.
658
+
659
+
660
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
661
+
662
+
663
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
664
+
665
+
666
+ Important considerations include:
667
+
668
+
669
+
670
+
671
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
672
+
673
+
674
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
675
+
676
+
677
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
678
+
679
+
680
+ Caution is needed when distantly related (≥3
681
+ rd
682
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
683
+ ACTC1 (HGNC:143),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
684
+ ACTC1 (HGNC:143),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
685
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
686
+ ACTC1 (HGNC:143),PP3,Supporting,"As many
687
+ in silico
688
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
689
+
690
+
691
+ Use of REVEL (Ioannidis
692
+ et al.
693
+ 2016
694
+ 12
695
+ ) is recommended at thresholds of
696
+ ≥0.70 for PP3
697
+ .
698
+
699
+
700
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
701
+
702
+
703
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
704
+
705
+
706
+ SpliceAI
707
+ 13
708
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
709
+ ACTC1 (HGNC:143),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
710
+ ACTC1 (HGNC:143),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
711
+ ACTC1 (HGNC:143),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
712
+ ACTC1 (HGNC:143),BA1,Stand Alone,"Allele frequency is
713
+ ≥0.001
714
+ based on the
715
+ filtering allele frequency (FAF)
716
+ in
717
+ gnomAD
718
+ in the subpopulation with the highest frequency (popmax).
719
+
720
+
721
+ The values used to calculate the BA1 threshold were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/300 or lower).
722
+
723
+
724
+ The threshold is applicable when assessing variants in the context of autosomal dominant cardiomyopathy. 
725
+
726
+
727
+ gnomAD is the preferred database for this calculation. If a subpopulation specific FAF other than the popmax is needed, this value can be calculated using the AlleleFrequencyApp on the
728
+ CardioDB website
729
+ .
730
+
731
+
732
+
733
+
734
+ Using the Inverse AF tab, enter in the population size and the number of alleles identified and it will calculate the FAF.  
735
+
736
+
737
+ Set confidence to 0.95 (95%).
738
+
739
+
740
+ If the FAF is ≥0.001, this rule can be applied.
741
+
742
+
743
+
744
+
745
+ The FAF by platform (e.g., exome vs. genome; v.2.1.1 vs. v.3.1.1) should be considered, the larger population is most likely to have the most accurate representation of “true” population allele frequency.
746
+
747
+
748
+ Caution is needed when considering any population cohorts that are smaller than the smallest subpopulations within gnomAD v.2.1.1 (e.g., ~5000 individuals or ~10,000 alleles). Despite this conservative nature of this threshold and approach, in smaller cohorts, the observed allele frequency may less accurately reflect the true allele frequency. Traditionally, once a variant is classified as Benign, it is rarely re-evaluated and so the highest confidence is needed to establish that classification on an allele frequency alone.",Disease-specific
749
+ ACTC1 (HGNC:143),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
750
+ ACTC1 (HGNC:143),BS1,Strong,"Allele frequency is
751
+ ≥0.0001 for
752
+
753
+ ACTC1
754
+ based on the
755
+ filtering allele frequency (FAF)
756
+ in
757
+ gnomAD
758
+ in the subpopulation with the highest frequency (popmax).
759
+
760
+
761
+ Criterion BS1 may only be used as standalone evidence to classify a variant as Likely Benign in the absence of conflicting data. See SVI guidance (Tavtigian
762
+ et al.
763
+ 2018
764
+ 14
765
+ ; Tavtigian
766
+ et al.
767
+ 2020
768
+ 15
769
+ ). 
770
+
771
+
772
+ See BA1 for additional specifications that also apply to BS1.","Disease-specific,Gene-specific"
773
+ ACTC1 (HGNC:143),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",NA
774
+ ACTC1 (HGNC:143),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
775
+ ACTC1 (HGNC:143),BS3,Strong,See PS3 specifications.,Disease-specific
776
+ ACTC1 (HGNC:143),BS3,Moderate,See PS3 specifications.,Disease-specific
777
+ ACTC1 (HGNC:143),BS3,Supporting,See PS3 specifications.,Disease-specific
778
+ ACTC1 (HGNC:143),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
779
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
780
+ ACTC1 (HGNC:143),BS4,Strong,"Any non-segregations should be carefully evaluated to rule out a phenocopy or the presence of a second disease-causing variant before considering it as conflicting or benign evidence. 
781
+
782
+
783
+
784
+
785
+ The presence of “phenocopies” (e.g., athlete’s heart, hypertensive heart disease, ischemic cardiomyopathy, alcoholic cardiomyopathy, diabetic cardiomyopathy) can mimic non-segregation (i.e., lack of segregation) among affected individuals. 
786
+
787
+
788
+ Families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent ‘non-segregation’.
789
+
790
+
791
+
792
+
793
+ Because of these possibilities,
794
+ multiple (≥2) non-segregations
795
+ that are highly unlikely to be phenocopies or due to alternate variants (e.g., those without a possible alternate cause)
796
+ are required to apply this rule
797
+ .  A higher number of non-segregations is necessary for instances where alternative causes are possible (e.g., non-segregation in a sibling with childhood onset cardiomyopathy versus a grandparent with hypertension and HCM).
798
+
799
+
800
+ Careful consideration of the above points is required when using this data as conflicting evidence, especially when overall evidence supports likely pathogenic or pathogenic.",Disease-specific
801
+ ACTC1 (HGNC:143),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
802
+ ACTC1 (HGNC:143),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
803
+ ACTC1 (HGNC:143),BP2,Supporting,"Other variants must be pathogenic as defined by these specifications.
804
+
805
+
806
+ Testing of parents or other informative relatives is often required to determine
807
+ cis
808
+ /
809
+ trans
810
+ status.
811
+
812
+
813
+ If a variant is seen in
814
+ trans
815
+ (or as double heterozygous) with another pathogenic variant in ≥2 cases and the phenotype is not more severe than when either of the two variants are seen in isolation, this rule may be applied (i.e., high confidence this variant is NOT contributing to disease).
816
+
817
+
818
+
819
+
820
+ <1% of cases of HCM have >1 pathogenic or likely pathogenic variant (0.6%; Alfares
821
+ et al.
822
+ 2015
823
+ 16
824
+ ).
825
+
826
+
827
+
828
+
829
+ This rule cannot be applied when the variant has only been observed in
830
+ cis
831
+ with a pathogenic variant as its significance in isolation is unknown in this scenario. 
832
+
833
+
834
+ Caution is needed if using this criterion as a primary piece of evidence for classifying a variant as likely benign/benign (i.e., only 2 SUPPORTING criteria are sufficient for a likely benign classification).",Disease-specific
835
+ ACTC1 (HGNC:143),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
836
+ ACTC1 (HGNC:143),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
837
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
838
+ ACTC1 (HGNC:143),BP4,Supporting,"As many
839
+ in silico
840
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
841
+
842
+
843
+ Use of REVEL (Ioannidis et al. 2016
844
+ 12
845
+ ) is recommended at thresholds of
846
+ ≤0.40 for BP4
847
+ .
848
+
849
+
850
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
851
+
852
+
853
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
854
+
855
+
856
+ SpliceAI
857
+ 13
858
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
859
+ ACTC1 (HGNC:143),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
860
+ ACTC1 (HGNC:143),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
861
+ ACTC1 (HGNC:143),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
862
+ ACTC1 (HGNC:143),BP7,Supporting,"Also applicable to
863
+ intronic variants outside the splice consensus sequence (-4 and +7 outward)
864
+ for which splicing prediction algorithms predict no impact to the splice consensus sequence NOR the creation of a new splice site AND the nucleotide is not highly conserved.
865
+
866
+
867
+ Rule can be combined with BP4 to make a variant likely benign per Richards
868
+ et al.
869
+ 2015
870
+ 1
871
+ .",General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYBPC3Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,952 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ MYBPC3 (HGNC:7551),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ MYBPC3 (HGNC:7551),PVS1,Very Strong,"Currently only applicable to
9
+ MYBPC3
10
+ where LOF is an established disease mechanism.
11
+
12
+
13
+ Refer to SVI guidance for the interpretation of this criterion (Abou Tayoun
14
+ et al.
15
+ 2018
16
+ 1
17
+ ).
18
+
19
+
20
+ SpliceAI
21
+ 2
22
+ is recommended for evaluation of predicted splice impacts.
23
+
24
+
25
+ Factors to consider when assessing the consequences of putative LOF variants in the
26
+ MYBPC3
27
+ gene:
28
+
29
+
30
+
31
+
32
+ Codon p.1254 is located 50 nucleotides upstream of the most 3' exon-exon junction (exon 33:34) in
33
+ MYBPC3.
34
+ As such, nonsense variants introducing a premature termination codon after this point may escape nonsense mediated decay (NMD) and consequently not result in protein haploinsufficiency (Nagy & Maquat 1998
35
+ 3
36
+ ).
37
+
38
+
39
+ When assessing variants predicted to affect splicing of micro-exons (exons 10, 11 and 14), be aware that
40
+ in silico
41
+ splice site predictions may be less reliable in this setting and the consequences of variants affecting splice sites at these exons less predictable (Frank-Hansen
42
+ et al.
43
+ 2008
44
+ 4
45
+ ).
46
+
47
+
48
+ When assessing variants affecting splice sites of in-frame exons (exons 2-4, 8-11, 14, 20, 22, 24-27), be aware that although most of these exons encode domains that have been shown to play critical roles in protein function, and/or harbor functionally important residues (Carrier
49
+ et al
50
+ . 2015
51
+ 5
52
+ ), in general, the consequences of in-frame deletions are less predictable.
53
+
54
+
55
+
56
+
57
+ For canonical splice site variants where other canonical splice variants have been reported, application of the PS1 rule may be considered if the other variant affecting the same slice site is 1) predicted to have a similar or more deleterious effect and 2) has been classified as pathogenic according to these modified guidelines without use of PS1 for other splice variants.
58
+
59
+
60
+ For genes where haploinsufficiency is NOT an established mechanism, see PM4 for truncating variants that do NOT undergo NMD.",Gene-specific
61
+ MYBPC3 (HGNC:7551),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
62
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
63
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
64
+ MYBPC3 (HGNC:7551),PS1,Strong,"No cardiomyopathy specifications. Apply as outlined by Richards
65
+ et al
66
+ . 2015
67
+ 6
68
+ .
69
+
70
+
71
+ Example of when rule should NOT be applied. NM_000256.3(
72
+ MYBPC3
73
+ ): c.2308G>A (p.Asp770Asn) has an established impact on splicing leading to nonsense mediated decay (NMD) and should not be used to provide evidence for other variants observed to result in the same amino acid change.",No change
74
+ MYBPC3 (HGNC:7551),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
75
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
76
+ MYBPC3 (HGNC:7551),PS2,Strong,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
77
+ 7
78
+ .
79
+
80
+
81
+ For most cardiomyopathies, it is recommended to default to
82
+ Phenotype consistency: “Phenotype consistent with gene but not highly specific”
83
+ . Clinical judgment is required for shifting to a higher or lower category. 
84
+
85
+
86
+ For use as a STRONG or VERY STRONG criterion, ideally parents have been thoroughly clinically evaluated without evidence of cardiomyopathy (ideally using a combination of ECG and echocardiogram or cardiac MRI for maximum sensitivity).
87
+
88
+
89
+ A family history consistent with
90
+ de novo
91
+ inheritance should not have any clinical signs or symptoms suggestive of cardiomyopathy in a 1
92
+ st
93
+ or 2
94
+ nd
95
+ degree relative, for example: 
96
+
97
+
98
+
99
+
100
+ Sudden death under 60 years of age
101
+
102
+
103
+ Heart transplant
104
+
105
+
106
+ Implantable cardiac defibrillator (ICD) under 60 years of age
107
+
108
+
109
+ Features of cardiomyopathy (e.g., systolic dysfunction, hypertrophy, left ventricular enlargement in an individual without risk factors).
110
+
111
+
112
+ Other related/overlapping cardiomyopathies
113
+
114
+
115
+
116
+
117
+ Examples of non-suspicious family history may include non-specific clinical features (e.g., palpitations, syncope, borderline/inconclusive echocardiogram findings, heart attack if age appropriate and suspected to result from coronary artery disease), but every attempt should be made to clarify features. 
118
+
119
+
120
+ Generally, this criterion is only applicable in the ABSENCE of any other possible disease-causing variants.  If other pathogenic or likely pathogenic variants are present, consider decreasing points assigned or overall weight.",Disease-specific
121
+ MYBPC3 (HGNC:7551),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
122
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
123
+ MYBPC3 (HGNC:7551),PS3,Strong,"In vitro splicing assays (e.g., RNA studies)
124
+
125
+
126
+ In vitro
127
+ splicing assays may be considered as 
128
+ STRONG
129
+ evidence, providing the following criteria are met.
130
+
131
+
132
+
133
+
134
+ Prior knowledge of predominant transcripts in cardiac tissue
135
+
136
+
137
+
138
+
139
+ Analysis undertaken using RNA extracted from cardiac tissue from the individual with the variant
140
+
141
+
142
+ Analysis undertaken using RNA extracted from whole blood providing the relevant transcripts (isoforms) are expressed in blood and are at sufficient levels to assess splice disruption.
143
+
144
+
145
+ Assay shows a clear, reproducible and convincing effect on splicing (i.e. a distinct splice product, present at a level comparable to the splice product from the wild-type allele), which is not observed in controls
146
+
147
+
148
+
149
+
150
+ Confirmation of abnormal splice product by Sanger sequencing
151
+
152
+
153
+
154
+
155
+ NOTE:
156
+ Mini-gene assay in non-patient derived cell lines are NOT considered to provide STRONG evidence.
157
+
158
+
159
+ NOTE:
160
+   Whether to activate this rule needs to be reconciled with the variant spectrum and disease mechanism for the gene at hand (i.e., consider whether the effect is likely to lead to LOF or an in-frame alteration and whether this type of effect is expected to be disease causing) (Abou Tayoun
161
+ et al.
162
+ 2018
163
+ 1
164
+ ).",Disease-specific
165
+ MYBPC3 (HGNC:7551),PS3,Moderate,"In vivo models (e.g., variant knock-in animal models)
166
+
167
+
168
+ Mammalian variant-specific knock-in animal models that produce a phenotype consistent with the clinical phenotype in humans (e.g., structural and/or functional cardiac abnormalities, premature death, arrhythmia) may be considered as
169
+ MODERATE
170
+ evidence
171
+
172
+
173
+ NOTE:
174
+ The following assays/models do NOT meet criteria
175
+
176
+
177
+
178
+
179
+ Assays that are known to be associated with non-specific cardiac phenotypes (e.g., morpholino-induced pericardial edema in zebrafish)
180
+
181
+
182
+ In vivo evidence that is not variant specific, such as whole gene alterations (i.e., cDNA or whole gene transgenic mice and whole or partial gene knock-out mice)",Disease-specific
183
+ MYBPC3 (HGNC:7551),PS3,Supporting,"In vitro
184
+
185
+ assays (e.g., biochemical assays of myofilament function, motility assays, human iPSC-CM)
186
+
187
+
188
+ While some
189
+ in vitro
190
+ assays may provide evidence that a variant in a cardiomyopathy gene has an effect on protein and/or myofilament function, at present, there are no validated “gold-standard” assays that are considered to reliably predict the clinical phenotype.
191
+
192
+
193
+ As such, in the cardiomyopathy genes listed in these guidelines, data from individual
194
+ in vitro
195
+ studies are unlikely to meet the criteria required to assign this rule at more than SUPPORTING level.",Disease-specific
196
+ MYBPC3 (HGNC:7551),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
197
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
198
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
199
+ MYBPC3 (HGNC:7551),PS4,Strong,"{Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
200
+
201
+
202
+ Cohorts used in these analyses should meet the following criteria: 
203
+
204
+
205
+
206
+
207
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
208
+
209
+
210
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
211
+
212
+
213
+
214
+
215
+
216
+
217
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
218
+
219
+
220
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
221
+
222
+
223
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
224
+
225
+
226
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
227
+
228
+
229
+ The population diversity of the case and control cohorts are broadly similar.
230
+
231
+
232
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
233
+ et al.
234
+ 2017
235
+ 9
236
+ ,
237
+ DECIPHER
238
+ ).
239
+
240
+
241
+
242
+
243
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
244
+
245
+
246
+
247
+
248
+ STRONG
249
+ evidence requires the lower bound of the 95% confidence interval (CI) around the odds ratio (OR) estimate to be
250
+ ≥20
251
+
252
+
253
+
254
+
255
+ A PS4 calculator is available at
256
+ www.cardiodb.org
257
+ .
258
+
259
+
260
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
261
+
262
+
263
+ *RELEVANT PHENOTYPES:
264
+
265
+
266
+
267
+
268
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
269
+
270
+
271
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
272
+
273
+
274
+ Additional considerations for LVNC and end-stage HCM: 
275
+
276
+
277
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
278
+ et al.
279
+ 2017
280
+ 10
281
+ ; Oechslin
282
+ et al.
283
+ 2017
284
+ 11
285
+ ; Hershberger
286
+ et al.
287
+ 2017
288
+ 12
289
+ ; Ross
290
+ et al.
291
+ 2020
292
+ 13
293
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
294
+
295
+
296
+
297
+
298
+
299
+
300
+
301
+
302
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
303
+ MYBPC3 (HGNC:7551),PS4,Moderate,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
304
+
305
+
306
+ Cohorts used in these analyses should meet the following criteria: 
307
+
308
+
309
+
310
+
311
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
312
+
313
+
314
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
315
+
316
+
317
+
318
+
319
+
320
+
321
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
322
+
323
+
324
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
325
+
326
+
327
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
328
+
329
+
330
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
331
+
332
+
333
+ The population diversity of the case and control cohorts are broadly similar.
334
+
335
+
336
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
337
+ et al.
338
+ 2017
339
+ 9
340
+ ,
341
+ DECIPHER
342
+ ).
343
+
344
+
345
+
346
+
347
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
348
+
349
+
350
+
351
+
352
+ MODERATE
353
+ evidence requires the lower bound of the 95% CI around the OR to be
354
+ ≥10
355
+
356
+
357
+
358
+
359
+ A PS4 calculator is available at
360
+ www.cardiodb.org
361
+ .
362
+
363
+
364
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
365
+
366
+
367
+ *RELEVANT PHENOTYPES:
368
+
369
+
370
+
371
+
372
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
373
+
374
+
375
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
376
+
377
+
378
+ Additional considerations for LVNC and end-stage HCM: 
379
+
380
+
381
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
382
+ et al.
383
+ 2017
384
+ 10
385
+ ; Oechslin
386
+ et al.
387
+ 2017
388
+ 11
389
+ ; Hershberger
390
+ et al.
391
+ 2017
392
+ 12
393
+ ; Ross
394
+ et al.
395
+ 2020
396
+ 13
397
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
398
+
399
+
400
+
401
+
402
+
403
+
404
+
405
+
406
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
407
+ MYBPC3 (HGNC:7551),PS4,Supporting,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
408
+
409
+
410
+ Cohorts used in these analyses should meet the following criteria: 
411
+
412
+
413
+
414
+
415
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
416
+
417
+
418
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
419
+
420
+
421
+
422
+
423
+
424
+
425
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
426
+
427
+
428
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
429
+
430
+
431
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
432
+
433
+
434
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
435
+
436
+
437
+ The population diversity of the case and control cohorts are broadly similar.
438
+
439
+
440
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
441
+ et al.
442
+ 2017
443
+ 9
444
+ ,
445
+ DECIPHER
446
+ ).
447
+
448
+
449
+
450
+
451
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
452
+
453
+
454
+
455
+
456
+ SUPPORTING
457
+ evidence requires the lower bound of the 95% CI around the OR to be
458
+ ≥5
459
+
460
+
461
+
462
+
463
+ A PS4 calculator is available at
464
+ www.cardiodb.org
465
+ .
466
+
467
+
468
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
469
+
470
+
471
+ *RELEVANT PHENOTYPES:
472
+
473
+
474
+
475
+
476
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
477
+
478
+
479
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
480
+
481
+
482
+ Additional considerations for LVNC and end-stage HCM: 
483
+
484
+
485
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
486
+ et al.
487
+ 2017
488
+ 10
489
+ ; Oechslin
490
+ et al.
491
+ 2017
492
+ 11
493
+ ; Hershberger
494
+ et al.
495
+ 2017
496
+ 12
497
+ ; Ross
498
+ et al.
499
+ 2020
500
+ 13
501
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
502
+
503
+
504
+
505
+
506
+
507
+
508
+
509
+
510
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
511
+ MYBPC3 (HGNC:7551),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
512
+ MYBPC3 (HGNC:7551),PM1,Moderate,"Applicable to missense variants in
513
+ MYBPC3
514
+ in the specific regions listed below (Walsh 
515
+ et al.
516
+ 2019
517
+ 14
518
+ ). 
519
+
520
+
521
+
522
+
523
+ Transcripts ENST00000545968 and NM_000256.3
524
+
525
+
526
+ Codons 485-502 and 1248-1266
527
+
528
+
529
+
530
+
531
+ Data from HCM case cohorts was used to derive these cluster regions. Therefore, this rule should NOT be applied when additional evidence for the variant supports that the variant causes a phenotype other than HCM (e.g., variant seen in multiple DCM cases).
532
+
533
+
534
+ Enrichment was not observed for DCM in any genes.
535
+
536
+
537
+ Rule should NOT be combined with PM5 because presence of pathogenic variants in the same codon/region were used to determine clustering and would be double-counting evidence.","Disease-specific,Gene-specific"
538
+ MYBPC3 (HGNC:7551),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
539
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
540
+ MYBPC3 (HGNC:7551),PM2,Supporting,"The values used to calculate the PM2 thresholds were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/500 or lower), where the most frequent pathogenic variant accounts for no more than 2% of cases (e.g., has an allele frequency of ≤0.02 in cases based on the upper bound of 95% CI), and where the penetrance of a pathogenic variant is expected to be at least 50% (Kelly
541
+ et al.
542
+ 2018
543
+ 15
544
+ ).
545
+
546
+
547
+ A threshold of
548
+ ≤0.00004
549
+ in the subpopulation with the highest frequency when using the upper bound of the 95% CI activates this rule.
550
+
551
+
552
+
553
+
554
+ Alternatively, this is equivalent to the variant NOT being observed more than once (≤1 allele) in gnomAD v.2.1.1 in one of the non-founder populations (e.g., absence required from the Other and Ashkenazi Jewish subpopulations).
555
+
556
+
557
+ Applying a threshold of ≤0.00004 (upper bound of 95% CI of the allele frequency in gnomAD) is equivalent to the variant being seen in a single subpopulation and that subpopulation meets any of the following:
558
+
559
+
560
+ Allele Count (AC) in Allele Number (AN)
561
+
562
+
563
+ ≤1 in ≥120,000
564
+
565
+
566
+ ≤2 in ≥160,000
567
+
568
+
569
+ ≤3 in ≥195,000
570
+
571
+
572
+ ≤4 in ≥230,000
573
+
574
+
575
+
576
+
577
+
578
+
579
+
580
+
581
+ gnomAD is the preferred database for this calculation, but currently only displays the filtering allele frequency (FAF), which is equivalent to a lower bound estimate of the 95% CI, when the upper bound is what is needed.
582
+
583
+
584
+
585
+
586
+ Confidence interval tools, such as
587
+ Confit-de-MAF
588
+ , can be used to determine the upper bound of the 95% CI of the observed allele frequency.
589
+
590
+
591
+
592
+
593
+ Due to current technical limitations of next generation sequencing technologies, minor allele frequencies for complex variants (e.g., large indels) may not be accurately represented in population databases.
594
+
595
+
596
+ Caution should be used when a variant is only identified, or over-represented, in one of the smaller gnomAD populations, as the gnomAD allele frequencies may not accurately represent the true population frequency.
597
+
598
+
599
+ Population databases may contain affected or pre-symptomatic individuals for diseases with reduced penetrance/variable onset.",Disease-specific
600
+ MYBPC3 (HGNC:7551),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
601
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
602
+ MYBPC3 (HGNC:7551),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
603
+ MYBPC3 (HGNC:7551),PM4,Moderate,"Strength of rule should be carefully considered and may require downgrading to SUPPORTING based on the predicted impact of the variant, including the size of the deletion/insertion, its location, and conservation of the region. 
604
+
605
+
606
+ For genes where PVS1 is not applicable (i.e., where there is no evidence that pLOF variants cause disease), consider using this rule at MODERATE or SUPPORTING strength for truncating variants that do NOT undergo nonsense mediated decay (NMD).",General recommendation
607
+ MYBPC3 (HGNC:7551),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
608
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
609
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
610
+ MYBPC3 (HGNC:7551),PM5,Moderate,"This criterion can be used at MODERATE if a different missense variant at the same codon has been classified as
611
+ pathogenic
612
+ using these modified guidelines without application of PM5.
613
+
614
+
615
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to SUPPORTING if the predicted impact is not expected to be equivalent or more severe.
616
+
617
+
618
+ PM5 should not be combined with PM1.  If both are applicable at MODERATE weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
619
+ MYBPC3 (HGNC:7551),PM5,Supporting,"This criterion can be considered at SUPPORTING if a different missense variant at the same codon has been classified as
620
+ likely pathogenic
621
+ using these modified guidelines without application of PM5.
622
+
623
+
624
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as likely pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to NOT APPLICABLE if the predicted impact is not expected to be equivalent or more severe.
625
+
626
+
627
+ PM5 should not be combined with PM1.  The one with the higher strength should be applied, but if both are applicable at SUPPORTING weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
628
+ MYBPC3 (HGNC:7551),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
629
+ MYBPC3 (HGNC:7551),PM6,Moderate,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
630
+ 7
631
+ .
632
+
633
+
634
+ For most cardiomyopathies, it is recommended to default to “phenotype consistent with gene but not highly specific”. Clinical judgment is required for shifting to a higher or lower phenotypic consistency. 
635
+
636
+
637
+ See PS2 for additional considerations.",Disease-specific
638
+ MYBPC3 (HGNC:7551),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
639
+ Note: May be used as stronger evidence with increasing segregation data.",
640
+ MYBPC3 (HGNC:7551),PP1,Strong,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
641
+ ≥7
642
+
643
+ segregations
644
+ (LOD score of 2.1) for
645
+ STRONG
646
+ .
647
+
648
+
649
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
650
+ 16
651
+ can be considered: 
652
+
653
+
654
+
655
+
656
+ STRONG evidence requires ≥5 segregations (LOD score of 1.5)
657
+
658
+
659
+
660
+
661
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
662
+
663
+
664
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
665
+
666
+
667
+ Important considerations include:
668
+
669
+
670
+
671
+
672
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
673
+
674
+
675
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1.
676
+
677
+
678
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
679
+
680
+
681
+ Caution is needed when distantly related (≥3
682
+ rd
683
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
684
+ MYBPC3 (HGNC:7551),PP1,Moderate,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
685
+ ≥5
686
+
687
+ segregations
688
+ (LOD score of 1.5) for
689
+ MODERATE
690
+ .
691
+
692
+
693
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
694
+ 16
695
+ can be considered: 
696
+
697
+
698
+
699
+
700
+ MODERATE evidence requires ≥4 segregations (LOD score of 1.2)
701
+
702
+
703
+
704
+
705
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
706
+
707
+
708
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
709
+
710
+
711
+ Important considerations include:
712
+
713
+
714
+
715
+
716
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
717
+
718
+
719
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
720
+
721
+
722
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
723
+
724
+
725
+ Caution is needed when distantly related (≥3
726
+ rd
727
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
728
+ MYBPC3 (HGNC:7551),PP1,Supporting,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
729
+ ≥3
730
+
731
+ segregations
732
+ (LOD score of 0.9) for
733
+ SUPPORTING
734
+ . The thresholds as proposed by Jarvik and Browning (2016)
735
+ 16
736
+ are the same at ≥3 segregations (LOD score of 0.9) for supporting.
737
+
738
+
739
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
740
+
741
+
742
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
743
+
744
+
745
+ Important considerations include:
746
+
747
+
748
+
749
+
750
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
751
+
752
+
753
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
754
+
755
+
756
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
757
+
758
+
759
+ Caution is needed when distantly related (≥3
760
+ rd
761
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
762
+ MYBPC3 (HGNC:7551),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
763
+ MYBPC3 (HGNC:7551),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
764
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
765
+ MYBPC3 (HGNC:7551),PP3,Supporting,"As many
766
+ in silico
767
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
768
+
769
+
770
+ Use of REVEL (Ioannidis
771
+ et al.
772
+ 2016
773
+ 17
774
+ ) is recommended at thresholds of
775
+ ≥0.70 for PP3
776
+ .
777
+
778
+
779
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
780
+
781
+
782
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
783
+
784
+
785
+ SpliceAI
786
+ 2
787
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
788
+ MYBPC3 (HGNC:7551),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
789
+ MYBPC3 (HGNC:7551),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
790
+ MYBPC3 (HGNC:7551),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
791
+ MYBPC3 (HGNC:7551),BA1,Stand Alone,"Allele frequency is
792
+ ≥0.001
793
+ based on the
794
+ filtering allele frequency (FAF)
795
+ in
796
+ gnomAD
797
+ in the subpopulation with the highest frequency (popmax).
798
+
799
+
800
+ The values used to calculate the BA1 threshold were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/300 or lower).
801
+
802
+
803
+ The threshold is applicable when assessing variants in the context of autosomal dominant cardiomyopathy. 
804
+
805
+
806
+ gnomAD is the preferred database for this calculation. If a subpopulation specific FAF other than the popmax is needed, this value can be calculated using the AlleleFrequencyApp on the
807
+ CardioDB website
808
+ .
809
+
810
+
811
+
812
+
813
+ Using the Inverse AF tab, enter in the population size and the number of alleles identified and it will calculate the FAF.  
814
+
815
+
816
+ Set confidence to 0.95 (95%).
817
+
818
+
819
+ If the FAF is ≥0.001, this rule can be applied.
820
+
821
+
822
+
823
+
824
+ The FAF by platform (e.g., exome vs. genome; v.2.1.1 vs. v.3.1.1) should be considered, the larger population is most likely to have the most accurate representation of “true” population allele frequency.
825
+
826
+
827
+ Caution is needed when considering any population cohorts that are smaller than the smallest subpopulations within gnomAD v.2.1.1 (e.g., ~5000 individuals or ~10,000 alleles). Despite this conservative nature of this threshold and approach, in smaller cohorts, the observed allele frequency may less accurately reflect the true allele frequency. Traditionally, once a variant is classified as Benign, it is rarely re-evaluated and so the highest confidence is needed to establish that classification on an allele frequency alone.",Disease-specific
828
+ MYBPC3 (HGNC:7551),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
829
+ MYBPC3 (HGNC:7551),BS1,Strong,"Allele frequency is
830
+ ≥0.0002 for
831
+
832
+ MYBPC3
833
+ based on the
834
+ filtering allele frequency (FAF)
835
+ in
836
+ gnomAD
837
+ in the subpopulation with the highest frequency (popmax).
838
+
839
+
840
+ Criterion BS1 may only be used as standalone evidence to classify a variant as Likely Benign in the absence of conflicting data. See SVI guidance (Tavtigian
841
+ et al.
842
+ 2018
843
+ 18
844
+ ; Tavtigian
845
+ et al.
846
+ 2020
847
+ 19
848
+ ). 
849
+
850
+
851
+ See BA1 for additional specifications that also apply to BS1.","Disease-specific,Gene-specific"
852
+ MYBPC3 (HGNC:7551),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",NA
853
+ MYBPC3 (HGNC:7551),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
854
+ MYBPC3 (HGNC:7551),BS3,Strong,See PS3 specifications.,Disease-specific
855
+ MYBPC3 (HGNC:7551),BS3,Moderate,See PS3 specifications.,Disease-specific
856
+ MYBPC3 (HGNC:7551),BS3,Supporting,See PS3 specifications.,Disease-specific
857
+ MYBPC3 (HGNC:7551),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
858
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
859
+ MYBPC3 (HGNC:7551),BS4,Strong,"Any non-segregations should be carefully evaluated to rule out a phenocopy or the presence of a second disease-causing variant before considering it as conflicting or benign evidence. 
860
+
861
+
862
+
863
+
864
+ The presence of “phenocopies” (e.g., athlete’s heart, hypertensive heart disease, ischemic cardiomyopathy, alcoholic cardiomyopathy, diabetic cardiomyopathy) can mimic non-segregation (i.e., lack of segregation) among affected individuals. 
865
+
866
+
867
+ Families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent ‘non-segregation’.
868
+
869
+
870
+
871
+
872
+ Because of these possibilities,
873
+ multiple (≥2) non-segregations
874
+ that are highly unlikely to be phenocopies or due to alternate variants (e.g., those without a possible alternate cause)
875
+ are required to apply this rule
876
+ .  A higher number of non-segregations is necessary for instances where alternative causes are possible (e.g., non-segregation in a sibling with childhood onset cardiomyopathy versus a grandparent with hypertension and HCM).
877
+
878
+
879
+ Careful consideration of the above points is required when using this data as conflicting evidence, especially when overall evidence supports likely pathogenic or pathogenic.",Disease-specific
880
+ MYBPC3 (HGNC:7551),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
881
+ MYBPC3 (HGNC:7551),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
882
+ MYBPC3 (HGNC:7551),BP2,Supporting,"Other variants must be pathogenic as defined by these specifications.
883
+
884
+
885
+ Testing of parents or other informative relatives is often required to determine
886
+ cis
887
+ /
888
+ trans
889
+ status.
890
+
891
+
892
+ If a variant is seen in
893
+ trans
894
+ (or as double heterozygous) with another pathogenic variant in ≥2 cases and the phenotype is not more severe than when either of the two variants are seen in isolation, this rule may be applied (i.e., high confidence this variant is NOT contributing to disease).
895
+
896
+
897
+
898
+
899
+ <1% of cases of HCM have >1 pathogenic or likely pathogenic variant (0.6%; Alfares
900
+ et al.
901
+ 2015
902
+ 20
903
+ ).
904
+
905
+
906
+
907
+
908
+ This rule cannot be applied when the variant has only been observed in
909
+ cis
910
+ with a pathogenic variant as its significance in isolation is unknown in this scenario. 
911
+
912
+
913
+ Caution is needed if using this criterion as a primary piece of evidence for classifying a variant as likely benign/benign (i.e., only 2 SUPPORTING criteria are sufficient for a likely benign classification).",Disease-specific
914
+ MYBPC3 (HGNC:7551),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
915
+ MYBPC3 (HGNC:7551),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
916
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
917
+ MYBPC3 (HGNC:7551),BP4,Supporting,"As many
918
+ in silico
919
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
920
+
921
+
922
+ Use of REVEL (Ioannidis
923
+ et al.
924
+ 2016
925
+ 17
926
+ ) is recommended at thresholds of
927
+ ≤0.40 for BP4
928
+ .
929
+
930
+
931
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
932
+
933
+
934
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
935
+
936
+
937
+ SpliceAI
938
+ 2
939
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
940
+ MYBPC3 (HGNC:7551),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
941
+ MYBPC3 (HGNC:7551),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
942
+ MYBPC3 (HGNC:7551),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
943
+ MYBPC3 (HGNC:7551),BP7,Supporting,"Also applicable to
944
+ intronic variants outside the splice consensus sequence (-4 and +7 outward)
945
+ for which splicing prediction algorithms predict no impact to the splice consensus sequence NOR the creation of a new splice site AND the nucleotide is not highly conserved.
946
+
947
+
948
+ Rule can be combined with BP4 to make a variant likely benign per Richards
949
+ et al.
950
+ 2015
951
+ 6
952
+ .",General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYH7Version2.0.0_version=2.0.0.csv ADDED
@@ -0,0 +1,900 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ MYH7 (HGNC:7577),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",NA
8
+ MYH7 (HGNC:7577),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
9
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
10
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
11
+ MYH7 (HGNC:7577),PS1,Strong,"No cardiomyopathy specifications. Apply as outlined by Richards
12
+ et al
13
+ . 2015
14
+ 1
15
+ .
16
+
17
+
18
+ Example of when rule should NOT be applied. NM_000256.3(
19
+ MYBPC3
20
+ ): c.2308G>A (p.Asp770Asn) has an established impact on splicing leading to nonsense mediated decay (NMD) and should not be used to provide evidence for other variants observed to result in the same amino acid change.",No change
21
+ MYH7 (HGNC:7577),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
22
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
23
+ MYH7 (HGNC:7577),PS2,Strong,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
24
+ 2
25
+ ).
26
+
27
+
28
+ For most cardiomyopathies, it is recommended to default to
29
+ Phenotype consistency: “Phenotype consistent with gene but not highly specific”
30
+ . Clinical judgment is required for shifting to a higher or lower category. 
31
+
32
+
33
+ For use as a STRONG or VERY STRONG criterion, ideally parents have been thoroughly clinically evaluated without evidence of cardiomyopathy (ideally using a combination of ECG and echocardiogram or cardiac MRI for maximum sensitivity).
34
+
35
+
36
+ A family history consistent with
37
+ de novo
38
+ inheritance should not have any clinical signs or symptoms suggestive of cardiomyopathy in a 1
39
+ st
40
+ or 2
41
+ nd
42
+ degree relative, for example: 
43
+
44
+
45
+
46
+
47
+ Sudden death under 60 years of age
48
+
49
+
50
+ Heart transplant
51
+
52
+
53
+ Implantable cardiac defibrillator (ICD) under 60 years of age
54
+
55
+
56
+ Features of cardiomyopathy (e.g., systolic dysfunction, hypertrophy, left ventricular enlargement in an individual without risk factors).
57
+
58
+
59
+ Other related/overlapping cardiomyopathies
60
+
61
+
62
+
63
+
64
+ Examples of non-suspicious family history may include non-specific clinical features (e.g., palpitations, syncope, borderline/inconclusive echocardiogram findings, heart attack if age appropriate and suspected to result from coronary artery disease), but every attempt should be made to clarify features. 
65
+
66
+
67
+ Generally, this criterion is only applicable in the ABSENCE of any other possible disease-causing variants.  If other pathogenic or likely pathogenic variants are present, consider decreasing points assigned or overall weight.",Disease-specific
68
+ MYH7 (HGNC:7577),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
69
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
70
+ MYH7 (HGNC:7577),PS3,Strong,"In vitro splicing assays (e.g., RNA studies)
71
+
72
+
73
+ In vitro
74
+ splicing assays may be considered as 
75
+ STRONG
76
+ evidence, providing the following criteria are met.
77
+
78
+
79
+
80
+
81
+ Prior knowledge of predominant transcripts in cardiac tissue
82
+
83
+
84
+
85
+
86
+ Analysis undertaken using RNA extracted from cardiac tissue from the individual with the variant
87
+
88
+
89
+ Analysis undertaken using RNA extracted from whole blood providing the relevant transcripts (isoforms) are expressed in blood and are at sufficient levels to assess splice disruption.
90
+
91
+
92
+ Assay shows a clear, reproducible and convincing effect on splicing (i.e. a distinct splice product, present at a level comparable to the splice product from the wild-type allele), which is not observed in controls
93
+
94
+
95
+
96
+
97
+ Confirmation of abnormal splice product by Sanger sequencing
98
+
99
+
100
+
101
+
102
+ NOTE:
103
+ Mini-gene assay in non-patient derived cell lines are NOT considered to provide STRONG evidence.
104
+
105
+
106
+ NOTE:
107
+   Whether to activate this rule needs to be reconciled with the variant spectrum and disease mechanism for the gene at hand (i.e., consider whether the effect is likely to lead to LOF or an in-frame alteration and whether this type of effect is expected to be disease causing) (Abou Tayoun
108
+ et al.
109
+ 2018
110
+ 3
111
+ ).",Disease-specific
112
+ MYH7 (HGNC:7577),PS3,Moderate,"In vivo models (e.g., variant knock-in animal models)
113
+
114
+
115
+ Mammalian variant-specific knock-in animal models that produce a phenotype consistent with the clinical phenotype in humans (e.g., structural and/or functional cardiac abnormalities, premature death, arrhythmia) may be considered as
116
+ MODERATE
117
+ evidence
118
+
119
+
120
+ NOTE:
121
+ The following assays/models do NOT meet criteria
122
+
123
+
124
+
125
+
126
+ Assays that are known to be associated with non-specific cardiac phenotypes (e.g., morpholino-induced pericardial edema in zebrafish)
127
+
128
+
129
+ In vivo evidence that is not variant specific, such as whole gene alterations (i.e., cDNA or whole gene transgenic mice and whole or partial gene knock-out mice)",Disease-specific
130
+ MYH7 (HGNC:7577),PS3,Supporting,"In vitro
131
+
132
+ assays (e.g., biochemical assays of myofilament function, motility assays, human iPSC-CM)
133
+
134
+
135
+ While some
136
+ in vitro
137
+ assays may provide evidence that a variant in a cardiomyopathy gene has an effect on protein and/or myofilament function, at present, there are no validated “gold-standard” assays that are considered to reliably predict the clinical phenotype.
138
+
139
+
140
+ As such, in the cardiomyopathy genes listed in these guidelines, data from individual
141
+ in vitro
142
+ studies are unlikely to meet the criteria required to assign this rule at more than SUPPORTING level.",Disease-specific
143
+ MYH7 (HGNC:7577),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
144
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
145
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
146
+ MYH7 (HGNC:7577),PS4,Strong,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
147
+
148
+
149
+ Cohorts used in these analyses should meet the following criteria: 
150
+
151
+
152
+
153
+
154
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
155
+
156
+
157
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
158
+
159
+
160
+
161
+
162
+
163
+
164
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
165
+
166
+
167
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
168
+
169
+
170
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
171
+
172
+
173
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
174
+
175
+
176
+ The population diversity of the case and control cohorts are broadly similar.
177
+
178
+
179
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
180
+ et al.
181
+ 2017
182
+ 8
183
+ ,
184
+ DECIPHER
185
+ ).
186
+
187
+
188
+
189
+
190
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
191
+
192
+
193
+
194
+
195
+ STRONG
196
+ evidence requires the lower bound of the 95% confidence interval (CI) around the odds ratio (OR) estimate to be
197
+ ≥20
198
+
199
+
200
+
201
+
202
+ A PS4 calculator is available at
203
+ www.cardiodb.org
204
+ .
205
+
206
+
207
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
208
+
209
+
210
+ *RELEVANT PHENOTYPES:
211
+
212
+
213
+
214
+
215
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
216
+
217
+
218
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
219
+
220
+
221
+ Additional considerations for LVNC and end-stage HCM: 
222
+
223
+
224
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
225
+ et al.
226
+ 2017
227
+ 9
228
+ ; Oechslin
229
+ et al.
230
+ 2017
231
+ 6
232
+ ; Hershberger
233
+ et al.
234
+ 2017
235
+ 5
236
+ ; Ross
237
+ et al.
238
+ 2020
239
+ 7
240
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
241
+
242
+
243
+
244
+
245
+
246
+
247
+
248
+
249
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
250
+ MYH7 (HGNC:7577),PS4,Moderate,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
251
+
252
+
253
+ Cohorts used in these analyses should meet the following criteria: 
254
+
255
+
256
+
257
+
258
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
259
+
260
+
261
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
262
+
263
+
264
+
265
+
266
+
267
+
268
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
269
+
270
+
271
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
272
+
273
+
274
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
275
+
276
+
277
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
278
+
279
+
280
+ The population diversity of the case and control cohorts are broadly similar.
281
+
282
+
283
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
284
+ et al.
285
+ 2017
286
+ 8
287
+ ,
288
+ DECIPHER
289
+ ).
290
+
291
+
292
+
293
+
294
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
295
+
296
+
297
+
298
+
299
+ MODERATE
300
+ evidence requires the lower bound of the 95% CI around the OR to be
301
+ ≥10
302
+
303
+
304
+
305
+
306
+ A PS4 calculator is available at
307
+ www.cardiodb.org
308
+ .
309
+
310
+
311
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
312
+
313
+
314
+ *RELEVANT PHENOTYPES:
315
+
316
+
317
+
318
+
319
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
320
+
321
+
322
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
323
+
324
+
325
+ Additional considerations for LVNC and end-stage HCM: 
326
+
327
+
328
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
329
+ et al.
330
+ 2017
331
+ 9
332
+ ; Oechslin
333
+ et al.
334
+ 2017
335
+ 6
336
+ ; Hershberger
337
+ et al.
338
+ 2017
339
+ 5
340
+ ; Ross
341
+ et al.
342
+ 2020
343
+ 7
344
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
345
+
346
+
347
+
348
+
349
+
350
+
351
+
352
+
353
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
354
+ MYH7 (HGNC:7577),PS4,Supporting,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
355
+
356
+
357
+ Cohorts used in these analyses should meet the following criteria: 
358
+
359
+
360
+
361
+
362
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
363
+
364
+
365
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
366
+
367
+
368
+
369
+
370
+
371
+
372
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
373
+
374
+
375
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
376
+
377
+
378
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
379
+
380
+
381
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
382
+
383
+
384
+ The population diversity of the case and control cohorts are broadly similar.
385
+
386
+
387
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
388
+ et al.
389
+ 2017
390
+ 8
391
+ ,
392
+ DECIPHER
393
+ ).
394
+
395
+
396
+
397
+
398
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
399
+
400
+
401
+
402
+
403
+ SUPPORTING
404
+ evidence requires the lower bound of the 95% CI around the OR to be
405
+ ≥5
406
+
407
+
408
+
409
+
410
+ A PS4 calculator is available at
411
+ www.cardiodb.org
412
+ .
413
+
414
+
415
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
416
+
417
+
418
+ *RELEVANT PHENOTYPES:
419
+
420
+
421
+
422
+
423
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
424
+
425
+
426
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
427
+
428
+
429
+ Additional considerations for LVNC and end-stage HCM: 
430
+
431
+
432
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
433
+ et al.
434
+ 2017
435
+ 9
436
+ ; Oechslin
437
+ et al.
438
+ 2017
439
+ 6
440
+ ; Hershberger
441
+ et al.
442
+ 2017
443
+ 5
444
+ ; Ross
445
+ et al.
446
+ 2020
447
+ 7
448
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
449
+
450
+
451
+
452
+
453
+
454
+
455
+
456
+
457
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
458
+ MYH7 (HGNC:7577),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
459
+ MYH7 (HGNC:7577),PM1,Moderate,"Applicable to missense variants in
460
+ MYH7
461
+ in the specific regions listed below (Walsh 
462
+ et al.
463
+ 2019
464
+ 10
465
+ ). 
466
+
467
+
468
+
469
+
470
+ Transcripts ENST00000355349 & NM_000257.4
471
+
472
+
473
+ Codons 167-931*
474
+
475
+
476
+
477
+
478
+ Data from HCM case cohorts was used to derive these cluster regions. Therefore, this rule should NOT be applied when additional evidence for the variant supports that the variant causes a phenotype other than HCM (e.g., variant seen in multiple DCM cases).
479
+
480
+
481
+ Enrichment was not observed for DCM in any genes.
482
+
483
+
484
+ Rule should NOT be combined with PM5 because presence of pathogenic variants in the same codon/region were used to determine clustering and would be double-counting evidence.
485
+
486
+
487
+ * This region is updated from v1.0 (Kelly et al. 2018
488
+ 11
489
+ ).","Disease-specific,Gene-specific"
490
+ MYH7 (HGNC:7577),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
491
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
492
+ MYH7 (HGNC:7577),PM2,Supporting,"The values used to calculate the PM2 thresholds were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/500 or lower), where the most frequent pathogenic variant accounts for no more than 2% of cases (e.g., has an allele frequency of ≤0.02 in cases based on the upper bound of 95% CI), and where the penetrance of a pathogenic variant is expected to be at least 50% (Kelly
493
+ et al.
494
+ 2018
495
+ 11
496
+ ).
497
+
498
+
499
+ A threshold of
500
+ ≤0.00004
501
+ in the subpopulation with the highest frequency when using the upper bound of the 95% CI activates this rule.
502
+
503
+
504
+
505
+
506
+ Alternatively, this is equivalent to the variant NOT being observed more than once (≤1 allele) in gnomAD v.2.1.1 in one of the non-founder populations (e.g., absence required from the Other and Ashkenazi Jewish subpopulations).
507
+
508
+
509
+ Applying a threshold of ≤0.00004 (upper bound of 95% CI of the allele frequency in gnomAD) is equivalent to the variant being seen in a single subpopulation and that subpopulation meets any of the following:
510
+
511
+
512
+ Allele Count (AC) in Allele Number (AN)
513
+
514
+
515
+ ≤1 in ≥120,000
516
+
517
+
518
+ ≤2 in ≥160,000
519
+
520
+
521
+ ≤3 in ≥195,000
522
+
523
+
524
+ ≤4 in ≥230,000
525
+
526
+
527
+
528
+
529
+
530
+
531
+
532
+
533
+ gnomAD is the preferred database for this calculation, but currently only displays the filtering allele frequency (FAF), which is equivalent to a lower bound estimate of the 95% CI, when the upper bound is what is needed.
534
+
535
+
536
+
537
+
538
+ Confidence interval tools, such as
539
+ Confit-de-MAF
540
+ , can be used to determine the upper bound of the 95% CI of the observed allele frequency.
541
+
542
+
543
+
544
+
545
+ Due to current technical limitations of next generation sequencing technologies, minor allele frequencies for complex variants (e.g., large indels) may not be accurately represented in population databases.
546
+
547
+
548
+ Caution should be used when a variant is only identified, or over-represented, in one of the smaller gnomAD populations, as the gnomAD allele frequencies may not accurately represent the true population frequency.
549
+
550
+
551
+ Population databases may contain affected or pre-symptomatic individuals for diseases with reduced penetrance/variable onset.",Disease-specific
552
+ MYH7 (HGNC:7577),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
553
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
554
+ MYH7 (HGNC:7577),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
555
+ MYH7 (HGNC:7577),PM4,Moderate,"Strength of rule should be carefully considered and may require downgrading to SUPPORTING based on the predicted impact of the variant, including the size of the deletion/insertion, its location, and conservation of the region. 
556
+
557
+
558
+ For genes where PVS1 is not applicable (i.e., where there is no evidence that pLOF variants cause disease), consider using this rule at MODERATE or SUPPORTING strength for truncating variants that do NOT undergo nonsense mediated decay (NMD).",Disease-specific
559
+ MYH7 (HGNC:7577),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
560
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
561
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
562
+ MYH7 (HGNC:7577),PM5,Moderate,"This criterion can be used at MODERATE if a different missense variant at the same codon has been classified as
563
+ pathogenic
564
+ using these modified guidelines without application of PM5.
565
+
566
+
567
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to SUPPORTING if the predicted impact is not expected to be equivalent or more severe.
568
+
569
+
570
+ PM5 should not be combined with PM1.  If both are applicable at MODERATE weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
571
+ MYH7 (HGNC:7577),PM5,Supporting,"This criterion can be considered at SUPPORTING if a different missense variant at the same codon has been classified as
572
+ likely pathogenic
573
+ using these modified guidelines without application of PM5.
574
+
575
+
576
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as likely pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to NOT APPLICABLE if the predicted impact is not expected to be equivalent or more severe.
577
+
578
+
579
+ PM5 should not be combined with PM1.  The one with the higher strength should be applied, but if both are applicable at SUPPORTING weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
580
+ MYH7 (HGNC:7577),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
581
+ MYH7 (HGNC:7577),PM6,Moderate,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
582
+ 2
583
+ .
584
+
585
+
586
+ For most cardiomyopathies, it is recommended to default to “phenotype consistent with gene but not highly specific”. Clinical judgment is required for shifting to a higher or lower phenotypic consistency. 
587
+
588
+
589
+ See PS2 for additional considerations.",Disease-specific
590
+ MYH7 (HGNC:7577),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
591
+ Note: May be used as stronger evidence with increasing segregation data.",
592
+ MYH7 (HGNC:7577),PP1,Strong,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
593
+ ≥7 segregations
594
+ (LOD score of 2.1) for
595
+ STRONG
596
+ .
597
+
598
+
599
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
600
+ 12
601
+ can be considered: 
602
+
603
+
604
+
605
+
606
+ STRONG evidence requires ≥5 segregations (LOD score of 1.5)
607
+
608
+
609
+
610
+
611
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
612
+
613
+
614
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
615
+
616
+
617
+ Important considerations include:
618
+
619
+
620
+
621
+
622
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
623
+
624
+
625
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1.
626
+
627
+
628
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
629
+
630
+
631
+ Caution is needed when distantly related (≥3
632
+ rd
633
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
634
+ MYH7 (HGNC:7577),PP1,Moderate,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
635
+ ≥5
636
+
637
+ segregations
638
+ (LOD score of 1.5) for
639
+ MODERATE
640
+ .
641
+
642
+
643
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
644
+ 12
645
+ can be considered: 
646
+
647
+
648
+
649
+
650
+ MODERATE evidence requires ≥4 segregations (LOD score of 1.2)
651
+
652
+
653
+
654
+
655
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
656
+
657
+
658
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
659
+
660
+
661
+ Important considerations include:
662
+
663
+
664
+
665
+
666
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
667
+
668
+
669
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
670
+
671
+
672
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
673
+
674
+
675
+ Caution is needed when distantly related (≥3
676
+ rd
677
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
678
+ MYH7 (HGNC:7577),PP1,Supporting,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
679
+ ≥3
680
+
681
+ segregations
682
+ (LOD score of 0.9) for
683
+ SUPPORTING
684
+ . The thresholds as proposed by Jarvik and Browning (2016)
685
+ 12
686
+ are the same at ≥3 segregations (LOD score of 0.9) for supporting.
687
+
688
+
689
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
690
+
691
+
692
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
693
+
694
+
695
+ Important considerations include:
696
+
697
+
698
+
699
+
700
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
701
+
702
+
703
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
704
+
705
+
706
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
707
+
708
+
709
+ Caution is needed when distantly related (≥3
710
+ rd
711
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
712
+ MYH7 (HGNC:7577),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
713
+ MYH7 (HGNC:7577),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
714
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
715
+ MYH7 (HGNC:7577),PP3,Supporting,"As many
716
+ in silico
717
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
718
+
719
+
720
+ Use of REVEL (Ioannidis
721
+ et al.
722
+ 2016
723
+ 14
724
+ ) is recommended at thresholds of
725
+ ≥0.70 for PP3
726
+ .
727
+
728
+
729
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
730
+
731
+
732
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
733
+
734
+
735
+ SpliceAI
736
+ 13
737
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
738
+ MYH7 (HGNC:7577),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
739
+ MYH7 (HGNC:7577),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
740
+ MYH7 (HGNC:7577),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
741
+ MYH7 (HGNC:7577),BA1,Stand Alone,"Allele frequency is
742
+ ≥0.001
743
+ based on the
744
+ filtering allele frequency (FAF)
745
+ in
746
+ gnomAD
747
+ in the subpopulation with the highest frequency (popmax).
748
+
749
+
750
+ The values used to calculate the BA1 threshold were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/300 or lower).
751
+
752
+
753
+ The threshold is applicable when assessing variants in the context of autosomal dominant cardiomyopathy. 
754
+
755
+
756
+ gnomAD is the preferred database for this calculation. If a subpopulation specific FAF other than the popmax is needed, this value can be calculated using the AlleleFrequencyApp on the
757
+ CardioDB website
758
+ .
759
+
760
+
761
+
762
+
763
+ Using the Inverse AF tab, enter in the population size and the number of alleles identified and it will calculate the FAF.  
764
+
765
+
766
+ Set confidence to 0.95 (95%).
767
+
768
+
769
+ If the FAF is ≥0.001, this rule can be applied.
770
+
771
+
772
+
773
+
774
+ The FAF by platform (e.g., exome vs. genome; v.2.1.1 vs. v.3.1.1) should be considered, the larger population is most likely to have the most accurate representation of “true” population allele frequency.
775
+
776
+
777
+ Caution is needed when considering any population cohorts that are smaller than the smallest subpopulations within gnomAD v.2.1.1 (e.g., ~5000 individuals or ~10,000 alleles). Despite this conservative nature of this threshold and approach, in smaller cohorts, the observed allele frequency may less accurately reflect the true allele frequency. Traditionally, once a variant is classified as Benign, it is rarely re-evaluated and so the highest confidence is needed to establish that classification on an allele frequency alone.",Disease-specific
778
+ MYH7 (HGNC:7577),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
779
+ MYH7 (HGNC:7577),BS1,Strong,"Allele frequency is
780
+ ≥0.0001 for
781
+
782
+ MYH7
783
+ based on the
784
+ filtering allele frequency (FAF)
785
+ in
786
+ gnomAD
787
+ in the subpopulation with the highest frequency (popmax).
788
+
789
+
790
+ Criterion BS1 may only be used as standalone evidence to classify a variant as Likely Benign in the absence of conflicting data. See SVI guidance (Tavtigian
791
+ et al.
792
+ 2018
793
+ 16
794
+ ; Tavtigian
795
+ et al.
796
+ 2020
797
+ 15
798
+ ). 
799
+
800
+
801
+ See BA1 for additional specifications that also apply to BS1.","Disease-specific,Gene-specific"
802
+ MYH7 (HGNC:7577),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",NA
803
+ MYH7 (HGNC:7577),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
804
+ MYH7 (HGNC:7577),BS3,Strong,See PS3 specifications.,No change
805
+ MYH7 (HGNC:7577),BS3,Moderate,See PS3 specifications.,Disease-specific
806
+ MYH7 (HGNC:7577),BS3,Supporting,See PS3 specifications.,Disease-specific
807
+ MYH7 (HGNC:7577),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
808
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
809
+ MYH7 (HGNC:7577),BS4,Strong,"Any non-segregations should be carefully evaluated to rule out a phenocopy or the presence of a second disease-causing variant before considering it as conflicting or benign evidence. 
810
+
811
+
812
+
813
+
814
+ The presence of “phenocopies” (e.g., athlete’s heart, hypertensive heart disease, ischemic cardiomyopathy, alcoholic cardiomyopathy, diabetic cardiomyopathy) can mimic non-segregation (i.e., lack of segregation) among affected individuals. 
815
+
816
+
817
+ Families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent ‘non-segregation’.
818
+
819
+
820
+
821
+
822
+ Because of these possibilities,
823
+ multiple (≥2) non-segregations
824
+ that are highly unlikely to be phenocopies or due to alternate variants (e.g., those without a possible alternate cause)
825
+ are required to apply this rule
826
+ .  A higher number of non-segregations is necessary for instances where alternative causes are possible (e.g., non-segregation in a sibling with childhood onset cardiomyopathy versus a grandparent with hypertension and HCM).
827
+
828
+
829
+ Careful consideration of the above points is required when using this data as conflicting evidence, especially when overall evidence supports likely pathogenic or pathogenic.",Disease-specific
830
+ MYH7 (HGNC:7577),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
831
+ MYH7 (HGNC:7577),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
832
+ MYH7 (HGNC:7577),BP2,Supporting,"Other variants must be pathogenic as defined by these specifications.
833
+
834
+
835
+ Testing of parents or other informative relatives is often required to determine
836
+ cis
837
+ /
838
+ trans
839
+ status.
840
+
841
+
842
+ If a variant is seen in
843
+ trans
844
+ (or as double heterozygous) with another pathogenic variant in ≥2 cases and the phenotype is not more severe than when either of the two variants are seen in isolation, this rule may be applied (i.e., high confidence this variant is NOT contributing to disease).
845
+
846
+
847
+
848
+
849
+ <1% of cases of HCM have >1 pathogenic or likely pathogenic variant (0.6%; Alfares
850
+ et al.
851
+ 2015
852
+ 17
853
+ ).
854
+
855
+
856
+
857
+
858
+ This rule cannot be applied when the variant has only been observed in
859
+ cis
860
+ with a pathogenic variant as its significance in isolation is unknown in this scenario. 
861
+
862
+
863
+ Caution is needed if using this criterion as a primary piece of evidence for classifying a variant as likely benign/benign (i.e., only 2 SUPPORTING criteria are sufficient for a likely benign classification).","Disease-specific,Gene-specific"
864
+ MYH7 (HGNC:7577),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
865
+ MYH7 (HGNC:7577),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
866
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
867
+ MYH7 (HGNC:7577),BP4,Supporting,"As many
868
+ in silico
869
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
870
+
871
+
872
+ Use of REVEL (Ioannidis et al. 2016
873
+ 14
874
+ ) is recommended at thresholds of
875
+ ≤0.40 for BP4
876
+ .
877
+
878
+
879
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
880
+
881
+
882
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
883
+
884
+
885
+ SpliceAI
886
+ 13
887
+ is recommended for evaluation of predicted splice impacts.",No change
888
+ MYH7 (HGNC:7577),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
889
+ MYH7 (HGNC:7577),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
890
+ MYH7 (HGNC:7577),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
891
+ MYH7 (HGNC:7577),BP7,Supporting,"Also applicable to
892
+ intronic variants outside the splice consensus sequence (-4 and +7 outward)
893
+ for which splicing prediction algorithms predict no impact to the splice consensus sequence NOR the creation of a new splice site AND the nucleotide is not highly conserved.
894
+
895
+
896
+ Rule can be combined with BP4 to make a variant likely benign per Richards
897
+ et al.
898
+ 2015
899
+ 1
900
+ .",General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYL2Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,882 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ MYL2 (HGNC:7583),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",NA
8
+ MYL2 (HGNC:7583),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
9
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
10
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
11
+ MYL2 (HGNC:7583),PS1,Strong,"No cardiomyopathy specifications. Apply as outlined by Richards
12
+ et al
13
+ . 2015
14
+ 1
15
+ .
16
+
17
+
18
+ Example of when rule should NOT be applied. NM_000256.3(
19
+ MYBPC3
20
+ ): c.2308G>A (p.Asp770Asn) has an established impact on splicing leading to nonsense mediated decay (NMD) and should not be used to provide evidence for other variants observed to result in the same amino acid change.",No change
21
+ MYL2 (HGNC:7583),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
22
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
23
+ MYL2 (HGNC:7583),PS2,Strong,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
24
+ 2
25
+ .
26
+
27
+
28
+ For most cardiomyopathies, it is recommended to default to
29
+ Phenotype consistency: “Phenotype consistent with gene but not highly specific”
30
+ . Clinical judgment is required for shifting to a higher or lower category. 
31
+
32
+
33
+ For use as a STRONG or VERY STRONG criterion, ideally parents have been thoroughly clinically evaluated without evidence of cardiomyopathy (ideally using a combination of ECG and echocardiogram or cardiac MRI for maximum sensitivity).
34
+
35
+
36
+ A family history consistent with
37
+ de novo
38
+ inheritance should not have any clinical signs or symptoms suggestive of cardiomyopathy in a 1
39
+ st
40
+ or 2
41
+ nd
42
+ degree relative, for example: 
43
+
44
+
45
+
46
+
47
+ Sudden death under 60 years of age
48
+
49
+
50
+ Heart transplant
51
+
52
+
53
+ Implantable cardiac defibrillator (ICD) under 60 years of age
54
+
55
+
56
+ Features of cardiomyopathy (e.g., systolic dysfunction, hypertrophy, left ventricular enlargement in an individual without risk factors).
57
+
58
+
59
+ Other related/overlapping cardiomyopathies
60
+
61
+
62
+
63
+
64
+ Examples of non-suspicious family history may include non-specific clinical features (e.g., palpitations, syncope, borderline/inconclusive echocardiogram findings, heart attack if age appropriate and suspected to result from coronary artery disease), but every attempt should be made to clarify features. 
65
+
66
+
67
+ Generally, this criterion is only applicable in the ABSENCE of any other possible disease-causing variants.  If other pathogenic or likely pathogenic variants are present, consider decreasing points assigned or overall weight.",Disease-specific
68
+ MYL2 (HGNC:7583),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
69
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
70
+ MYL2 (HGNC:7583),PS3,Strong,"In vitro splicing assays (e.g., RNA studies)
71
+
72
+
73
+ In vitro
74
+ splicing assays may be considered as 
75
+ STRONG
76
+ evidence, providing the following criteria are met.
77
+
78
+
79
+
80
+
81
+ Prior knowledge of predominant transcripts in cardiac tissue
82
+
83
+
84
+
85
+
86
+ Analysis undertaken using RNA extracted from cardiac tissue from the individual with the variant
87
+
88
+
89
+ Analysis undertaken using RNA extracted from whole blood providing the relevant transcripts (isoforms) are expressed in blood and are at sufficient levels to assess splice disruption.
90
+
91
+
92
+ Assay shows a clear, reproducible and convincing effect on splicing (i.e. a distinct splice product, present at a level comparable to the splice product from the wild-type allele), which is not observed in controls
93
+
94
+
95
+
96
+
97
+ Confirmation of abnormal splice product by Sanger sequencing
98
+
99
+
100
+
101
+
102
+ NOTE:
103
+ Mini-gene assay in non-patient derived cell lines are NOT considered to provide STRONG evidence.
104
+
105
+
106
+ NOTE:
107
+   Whether to activate this rule needs to be reconciled with the variant spectrum and disease mechanism for the gene at hand (i.e., consider whether the effect is likely to lead to LOF or an in-frame alteration and whether this type of effect is expected to be disease causing) (Abou Tayoun
108
+ et al.
109
+ 2018
110
+ 3
111
+ ).",Disease-specific
112
+ MYL2 (HGNC:7583),PS3,Moderate,"In vivo models (e.g., variant knock-in animal models)
113
+
114
+
115
+ Mammalian variant-specific knock-in animal models that produce a phenotype consistent with the clinical phenotype in humans (e.g., structural and/or functional cardiac abnormalities, premature death, arrhythmia) may be considered as
116
+ MODERATE
117
+ evidence
118
+
119
+
120
+ NOTE:
121
+ The following assays/models do NOT meet criteria
122
+
123
+
124
+
125
+
126
+ Assays that are known to be associated with non-specific cardiac phenotypes (e.g., morpholino-induced pericardial edema in zebrafish)
127
+
128
+
129
+ In vivo evidence that is not variant specific, such as whole gene alterations (i.e., cDNA or whole gene transgenic mice and whole or partial gene knock-out mice)",Disease-specific
130
+ MYL2 (HGNC:7583),PS3,Supporting,"In vitro
131
+
132
+ assays (e.g., biochemical assays of myofilament function, motility assays, human iPSC-CM)
133
+
134
+
135
+ While some
136
+ in vitro
137
+ assays may provide evidence that a variant in a cardiomyopathy gene has an effect on protein and/or myofilament function, at present, there are no validated “gold-standard” assays that are considered to reliably predict the clinical phenotype.
138
+
139
+
140
+ As such, in the cardiomyopathy genes listed in these guidelines, data from individual
141
+ in vitro
142
+ studies are unlikely to meet the criteria required to assign this rule at more than SUPPORTING level.",Disease-specific
143
+ MYL2 (HGNC:7583),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
144
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
145
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
146
+ MYL2 (HGNC:7583),PS4,Strong,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
147
+
148
+
149
+ Cohorts used in these analyses should meet the following criteria: 
150
+
151
+
152
+
153
+
154
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
155
+
156
+
157
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
158
+
159
+
160
+
161
+
162
+
163
+
164
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
165
+
166
+
167
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
168
+
169
+
170
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
171
+
172
+
173
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
174
+
175
+
176
+ The population diversity of the case and control cohorts are broadly similar.
177
+
178
+
179
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
180
+ et al.
181
+ 2017
182
+ 5
183
+ ,
184
+ DECIPHER
185
+ ).
186
+
187
+
188
+
189
+
190
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
191
+
192
+
193
+
194
+
195
+ STRONG
196
+ evidence requires the lower bound of the 95% confidence interval (CI) around the odds ratio (OR) estimate to be
197
+ ≥20
198
+
199
+
200
+
201
+
202
+ A PS4 calculator is available at
203
+ www.cardiodb.org
204
+ .
205
+
206
+
207
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
208
+
209
+
210
+ *RELEVANT PHENOTYPES:
211
+
212
+
213
+
214
+
215
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
216
+
217
+
218
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
219
+
220
+
221
+ Additional considerations for LVNC and end-stage HCM: 
222
+
223
+
224
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
225
+ et al.
226
+ 2017
227
+ 6
228
+ ; Oechslin
229
+ et al.
230
+ 2017
231
+ 7
232
+ ; Hershberger
233
+ et al.
234
+ 2017
235
+ 8
236
+ ; Ross
237
+ et al.
238
+ 2020
239
+ 9
240
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
241
+
242
+
243
+
244
+
245
+
246
+
247
+
248
+
249
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
250
+ MYL2 (HGNC:7583),PS4,Moderate,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
251
+
252
+
253
+ Cohorts used in these analyses should meet the following criteria: 
254
+
255
+
256
+
257
+
258
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
259
+
260
+
261
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
262
+
263
+
264
+
265
+
266
+
267
+
268
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
269
+
270
+
271
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
272
+
273
+
274
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
275
+
276
+
277
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
278
+
279
+
280
+ The population diversity of the case and control cohorts are broadly similar.
281
+
282
+
283
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
284
+ et al.
285
+ 2017
286
+ 5
287
+ ,
288
+ DECIPHER
289
+ ).
290
+
291
+
292
+
293
+
294
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
295
+
296
+
297
+
298
+
299
+ MODERATE
300
+ evidence requires the lower bound of the 95% CI around the OR to be
301
+ ≥10
302
+
303
+
304
+
305
+
306
+ A PS4 calculator is available at
307
+ www.cardiodb.org
308
+ .
309
+
310
+
311
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
312
+
313
+
314
+ *RELEVANT PHENOTYPES:
315
+
316
+
317
+
318
+
319
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
320
+
321
+
322
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
323
+
324
+
325
+ Additional considerations for LVNC and end-stage HCM: 
326
+
327
+
328
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
329
+ et al.
330
+ 2017
331
+ 6
332
+ ; Oechslin
333
+ et al.
334
+ 2017
335
+ 7
336
+ ; Hershberger
337
+ et al.
338
+ 2017
339
+ 8
340
+ ; Ross
341
+ et al.
342
+ 2020
343
+ 9
344
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
345
+
346
+
347
+
348
+
349
+
350
+
351
+
352
+
353
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
354
+ MYL2 (HGNC:7583),PS4,Supporting,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
355
+
356
+
357
+ Cohorts used in these analyses should meet the following criteria: 
358
+
359
+
360
+
361
+
362
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
363
+
364
+
365
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
366
+
367
+
368
+
369
+
370
+
371
+
372
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
373
+
374
+
375
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
376
+
377
+
378
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
379
+
380
+
381
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
382
+
383
+
384
+ The population diversity of the case and control cohorts are broadly similar.
385
+
386
+
387
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
388
+ et al.
389
+ 2017
390
+ 5
391
+ ,
392
+ DECIPHER
393
+ ).
394
+
395
+
396
+
397
+
398
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
399
+
400
+
401
+
402
+
403
+ SUPPORTING
404
+ evidence requires the lower bound of the 95% CI around the OR to be
405
+ ≥5
406
+
407
+
408
+
409
+
410
+ A PS4 calculator is available at
411
+ www.cardiodb.org
412
+ .
413
+
414
+
415
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
416
+
417
+
418
+ *RELEVANT PHENOTYPES:
419
+
420
+
421
+
422
+
423
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
424
+
425
+
426
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
427
+
428
+
429
+ Additional considerations for LVNC and end-stage HCM: 
430
+
431
+
432
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
433
+ et al.
434
+ 2017
435
+ 6
436
+ ; Oechslin
437
+ et al.
438
+ 2017
439
+ 7
440
+ ; Hershberger
441
+ et al.
442
+ 2017
443
+ 8
444
+ ; Ross
445
+ et al.
446
+ 2020
447
+ 9
448
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
449
+
450
+
451
+
452
+
453
+
454
+
455
+
456
+
457
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
458
+ MYL2 (HGNC:7583),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
459
+ MYL2 (HGNC:7583),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
460
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
461
+ MYL2 (HGNC:7583),PM2,Supporting,"The values used to calculate the PM2 thresholds were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/500 or lower), where the most frequent pathogenic variant accounts for no more than 2% of cases (e.g., has an allele frequency of ≤0.02 in cases based on the upper bound of 95% CI), and where the penetrance of a pathogenic variant is expected to be at least 50% (Kelly
462
+ et al.
463
+ 2018
464
+ 10
465
+ ).
466
+
467
+
468
+ A threshold of
469
+ ≤0.00004
470
+ in the subpopulation with the highest frequency when using the upper bound of the 95% CI activates this rule.
471
+
472
+
473
+
474
+
475
+ Alternatively, this is equivalent to the variant NOT being observed more than once (≤1 allele) in gnomAD v.2.1.1 in one of the non-founder populations (e.g., absence required from the Other and Ashkenazi Jewish subpopulations).
476
+
477
+
478
+ Applying a threshold of ≤0.00004 (upper bound of 95% CI of the allele frequency in gnomAD) is equivalent to the variant being seen in a single subpopulation and that subpopulation meets any of the following:
479
+
480
+
481
+ Allele Count (AC) in Allele Number (AN)
482
+
483
+
484
+ ≤1 in ≥120,000
485
+
486
+
487
+ ≤2 in ≥160,000
488
+
489
+
490
+ ≤3 in ≥195,000
491
+
492
+
493
+ ≤4 in ≥230,000
494
+
495
+
496
+
497
+
498
+
499
+
500
+
501
+
502
+ gnomAD is the preferred database for this calculation, but currently only displays the filtering allele frequency (FAF), which is equivalent to a lower bound estimate of the 95% CI, when the upper bound is what is needed.
503
+
504
+
505
+
506
+
507
+ Confidence interval tools, such as
508
+ Confit-de-MAF
509
+ , can be used to determine the upper bound of the 95% CI of the observed allele frequency.
510
+
511
+
512
+
513
+
514
+ Due to current technical limitations of next generation sequencing technologies, minor allele frequencies for complex variants (e.g., large indels) may not be accurately represented in population databases.
515
+
516
+
517
+ Caution should be used when a variant is only identified, or over-represented, in one of the smaller gnomAD populations, as the gnomAD allele frequencies may not accurately represent the true population frequency.
518
+
519
+
520
+ Population databases may contain affected or pre-symptomatic individuals for diseases with reduced penetrance/variable onset.",Disease-specific
521
+ MYL2 (HGNC:7583),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
522
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
523
+ MYL2 (HGNC:7583),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
524
+ MYL2 (HGNC:7583),PM4,Moderate,"Strength of rule should be carefully considered and may require downgrading to SUPPORTING based on the predicted impact of the variant, including the size of the deletion/insertion, its location, and conservation of the region. 
525
+
526
+
527
+ For genes where PVS1 is not applicable (i.e., where there is no evidence that pLOF variants cause disease), consider using this rule at MODERATE or SUPPORTING strength for truncating variants that do NOT undergo nonsense mediated decay (NMD).",General recommendation
528
+ MYL2 (HGNC:7583),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
529
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
530
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
531
+ MYL2 (HGNC:7583),PM5,Moderate,"This criterion can be used at MODERATE if a different missense variant at the same codon has been classified as
532
+ pathogenic
533
+ using these modified guidelines without application of PM5.
534
+
535
+
536
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to SUPPORTING if the predicted impact is not expected to be equivalent or more severe.
537
+
538
+
539
+ PM5 should not be combined with PM1.  If both are applicable at MODERATE weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
540
+ MYL2 (HGNC:7583),PM5,Supporting,"This criterion can be considered at SUPPORTING if a different missense variant at the same codon has been classified as
541
+ likely pathogenic
542
+ using these modified guidelines without application of PM5.
543
+
544
+
545
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as likely pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to NOT APPLICABLE if the predicted impact is not expected to be equivalent or more severe.
546
+
547
+
548
+ PM5 should not be combined with PM1.  The one with the higher strength should be applied, but if both are applicable at SUPPORTING weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
549
+ MYL2 (HGNC:7583),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
550
+ MYL2 (HGNC:7583),PM6,Moderate,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
551
+ 2
552
+ .
553
+
554
+
555
+ For most cardiomyopathies, it is recommended to default to “phenotype consistent with gene but not highly specific”. Clinical judgment is required for shifting to a higher or lower phenotypic consistency. 
556
+
557
+
558
+ See PS2 for additional considerations.",Disease-specific
559
+ MYL2 (HGNC:7583),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
560
+ Note: May be used as stronger evidence with increasing segregation data.",
561
+ MYL2 (HGNC:7583),PP1,Strong,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
562
+ ≥7
563
+
564
+ segregations
565
+ (LOD score of 2.1) for
566
+ STRONG
567
+ .
568
+
569
+
570
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
571
+ 11
572
+ can be considered: 
573
+
574
+
575
+
576
+
577
+ STRONG evidence requires ≥5 segregations (LOD score of 1.5)
578
+
579
+
580
+
581
+
582
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
583
+
584
+
585
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
586
+
587
+
588
+ Important considerations include:
589
+
590
+
591
+
592
+
593
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
594
+
595
+
596
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1.
597
+
598
+
599
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
600
+
601
+
602
+ Caution is needed when distantly related (≥3
603
+ rd
604
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
605
+ MYL2 (HGNC:7583),PP1,Moderate,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
606
+ ≥5
607
+
608
+ segregations
609
+ (LOD score of 1.5) for
610
+ MODERATE
611
+ .
612
+
613
+
614
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
615
+ 11
616
+ can be considered: 
617
+
618
+
619
+
620
+
621
+ MODERATE evidence requires ≥4 segregations (LOD score of 1.2)
622
+
623
+
624
+
625
+
626
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
627
+
628
+
629
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
630
+
631
+
632
+ Important considerations include:
633
+
634
+
635
+
636
+
637
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
638
+
639
+
640
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
641
+
642
+
643
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
644
+
645
+
646
+ Caution is needed when distantly related (≥3
647
+ rd
648
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
649
+ MYL2 (HGNC:7583),PP1,Supporting,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
650
+ ≥3
651
+
652
+ segregations
653
+ (LOD score of 0.9) for
654
+ SUPPORTING
655
+ . The thresholds as proposed by Jarvik and Browning (2016)
656
+ 11
657
+ are the same at ≥3 segregations (LOD score of 0.9) for supporting.
658
+
659
+
660
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
661
+
662
+
663
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
664
+
665
+
666
+ Important considerations include:
667
+
668
+
669
+
670
+
671
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
672
+
673
+
674
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
675
+
676
+
677
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
678
+
679
+
680
+ Caution is needed when distantly related (≥3
681
+ rd
682
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
683
+ MYL2 (HGNC:7583),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
684
+ MYL2 (HGNC:7583),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
685
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
686
+ MYL2 (HGNC:7583),PP3,Supporting,"As many
687
+ in silico
688
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
689
+
690
+
691
+ Use of REVEL (Ioannidis
692
+ et al.
693
+ 2016
694
+ 12
695
+ ) is recommended at thresholds of
696
+ ≥0.70 for PP3
697
+ .
698
+
699
+
700
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
701
+
702
+
703
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
704
+
705
+
706
+ SpliceAI
707
+ 13
708
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
709
+ MYL2 (HGNC:7583),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
710
+ MYL2 (HGNC:7583),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
711
+ MYL2 (HGNC:7583),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
712
+ MYL2 (HGNC:7583),BA1,Stand Alone,"Allele frequency is
713
+ ≥0.001
714
+ based on the
715
+ filtering allele frequency (FAF)
716
+ in
717
+ gnomAD
718
+ in the subpopulation with the highest frequency (popmax).
719
+
720
+
721
+ The values used to calculate the BA1 threshold were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/300 or lower).
722
+
723
+
724
+ The threshold is applicable when assessing variants in the context of autosomal dominant cardiomyopathy. 
725
+
726
+
727
+
728
+
729
+ Caution should be applied when assessing variants in the
730
+ MYL2
731
+ and
732
+ MYL3
733
+ genes, as homozygous or compound heterozygous variants have been reported to cause a recessive HCM and heterozygous individuals show no sign of disease.
734
+
735
+
736
+
737
+
738
+ gnomAD is the preferred database for this calculation. If a subpopulation specific FAF other than the popmax is needed, this value can be calculated using the AlleleFrequencyApp on the
739
+ CardioDB website
740
+ .
741
+
742
+
743
+
744
+
745
+ Using the Inverse AF tab, enter in the population size and the number of alleles identified and it will calculate the FAF.  
746
+
747
+
748
+ Set confidence to 0.95 (95%).
749
+
750
+
751
+ If the FAF is ≥0.001, this rule can be applied.
752
+
753
+
754
+
755
+
756
+ The FAF by platform (e.g., exome vs. genome; v.2.1.1 vs. v.3.1.1) should be considered, the larger population is most likely to have the most accurate representation of “true” population allele frequency.
757
+
758
+
759
+ Caution is needed when considering any population cohorts that are smaller than the smallest subpopulations within gnomAD v.2.1.1 (e.g., ~5000 individuals or ~10,000 alleles). Despite this conservative nature of this threshold and approach, in smaller cohorts, the observed allele frequency may less accurately reflect the true allele frequency. Traditionally, once a variant is classified as Benign, it is rarely re-evaluated and so the highest confidence is needed to establish that classification on an allele frequency alone.",Disease-specific
760
+ MYL2 (HGNC:7583),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
761
+ MYL2 (HGNC:7583),BS1,Strong,"Allele frequency is
762
+ ≥0.0001 for
763
+
764
+ MYL2
765
+ based on the
766
+ filtering allele frequency (FAF)
767
+ in
768
+ gnomAD
769
+ in the subpopulation with the highest frequency (popmax).
770
+
771
+
772
+ Criterion BS1 may only be used as standalone evidence to classify a variant as Likely Benign in the absence of conflicting data. See SVI guidance (Tavtigian
773
+ et al.
774
+ 2018
775
+ 14
776
+ ; Tavtigian
777
+ et al.
778
+ 2020
779
+ 15
780
+ ). 
781
+
782
+
783
+ See BA1 for additional specifications that also apply to BS1.","Disease-specific,Gene-specific"
784
+ MYL2 (HGNC:7583),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",NA
785
+ MYL2 (HGNC:7583),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
786
+ MYL2 (HGNC:7583),BS3,Strong,See PS3 specifications.,Disease-specific
787
+ MYL2 (HGNC:7583),BS3,Moderate,See PS3 specifications.,Disease-specific
788
+ MYL2 (HGNC:7583),BS3,Supporting,See PS3 specifications.,Disease-specific
789
+ MYL2 (HGNC:7583),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
790
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
791
+ MYL2 (HGNC:7583),BS4,Strong,"Any non-segregations should be carefully evaluated to rule out a phenocopy or the presence of a second disease-causing variant before considering it as conflicting or benign evidence. 
792
+
793
+
794
+
795
+
796
+ The presence of “phenocopies” (e.g., athlete’s heart, hypertensive heart disease, ischemic cardiomyopathy, alcoholic cardiomyopathy, diabetic cardiomyopathy) can mimic non-segregation (i.e., lack of segregation) among affected individuals. 
797
+
798
+
799
+ Families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent ‘non-segregation’.
800
+
801
+
802
+
803
+
804
+ Because of these possibilities,
805
+ multiple (≥2) non-segregations
806
+ that are highly unlikely to be phenocopies or due to alternate variants (e.g., those without a possible alternate cause)
807
+ are required to apply this rule
808
+ .  A higher number of non-segregations is necessary for instances where alternative causes are possible (e.g., non-segregation in a sibling with childhood onset cardiomyopathy versus a grandparent with hypertension and HCM).
809
+
810
+
811
+ Careful consideration of the above points is required when using this data as conflicting evidence, especially when overall evidence supports likely pathogenic or pathogenic.",Disease-specific
812
+ MYL2 (HGNC:7583),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
813
+ MYL2 (HGNC:7583),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
814
+ MYL2 (HGNC:7583),BP2,Supporting,"Other variants must be pathogenic as defined by these specifications.
815
+
816
+
817
+ Testing of parents or other informative relatives is often required to determine
818
+ cis
819
+ /
820
+ trans
821
+ status.
822
+
823
+
824
+ If a variant is seen in
825
+ trans
826
+ (or as double heterozygous) with another pathogenic variant in ≥2 cases and the phenotype is not more severe than when either of the two variants are seen in isolation, this rule may be applied (i.e., high confidence this variant is NOT contributing to disease).
827
+
828
+
829
+
830
+
831
+ <1% of cases of HCM have >1 pathogenic or likely pathogenic variant (0.6%; Alfares
832
+ et al.
833
+ 2015
834
+ 16
835
+ ).
836
+
837
+
838
+
839
+
840
+ This rule cannot be applied when the variant has only been observed in
841
+ cis
842
+ with a pathogenic variant as its significance in isolation is unknown in this scenario. 
843
+
844
+
845
+ Caution is needed if using this criterion as a primary piece of evidence for classifying a variant as likely benign/benign (i.e., only 2 SUPPORTING criteria are sufficient for a likely benign classification).",Disease-specific
846
+ MYL2 (HGNC:7583),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
847
+ MYL2 (HGNC:7583),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
848
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
849
+ MYL2 (HGNC:7583),BP4,Supporting,"As many
850
+ in silico
851
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
852
+
853
+
854
+ Use of REVEL (Ioannidis et al. 2016
855
+ 12
856
+ ) is recommended at thresholds of
857
+ ≤0.40 for BP4
858
+ .
859
+
860
+
861
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
862
+
863
+
864
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
865
+
866
+
867
+ SpliceAI
868
+ 13
869
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
870
+ MYL2 (HGNC:7583),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
871
+ MYL2 (HGNC:7583),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
872
+ MYL2 (HGNC:7583),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
873
+ MYL2 (HGNC:7583),BP7,Supporting,"Also applicable to
874
+ intronic variants outside the splice consensus sequence (-4 and +7 outward)
875
+ for which splicing prediction algorithms predict no impact to the splice consensus sequence NOR the creation of a new splice site AND the nucleotide is not highly conserved.
876
+
877
+
878
+ Rule can be combined with BP4 to make a variant likely benign per Richards
879
+ et al.
880
+ 2015
881
+ 1
882
+ .",General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMYL3Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,882 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ MYL3 (HGNC:7584),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",NA
8
+ MYL3 (HGNC:7584),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
9
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
10
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
11
+ MYL3 (HGNC:7584),PS1,Strong,"No cardiomyopathy specifications. Apply as outlined by Richards
12
+ et al
13
+ . 2015
14
+ 1
15
+ .
16
+
17
+
18
+ Example of when rule should NOT be applied. NM_000256.3(
19
+ MYBPC3
20
+ ): c.2308G>A (p.Asp770Asn) has an established impact on splicing leading to nonsense mediated decay (NMD) and should not be used to provide evidence for other variants observed to result in the same amino acid change.",No change
21
+ MYL3 (HGNC:7584),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
22
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
23
+ MYL3 (HGNC:7584),PS2,Strong,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
24
+ 2
25
+ .
26
+
27
+
28
+ For most cardiomyopathies, it is recommended to default to
29
+ Phenotype consistency: “Phenotype consistent with gene but not highly specific”
30
+ . Clinical judgment is required for shifting to a higher or lower category. 
31
+
32
+
33
+ For use as a STRONG or VERY STRONG criterion, ideally parents have been thoroughly clinically evaluated without evidence of cardiomyopathy (ideally using a combination of ECG and echocardiogram or cardiac MRI for maximum sensitivity).
34
+
35
+
36
+ A family history consistent with
37
+ de novo
38
+ inheritance should not have any clinical signs or symptoms suggestive of cardiomyopathy in a 1
39
+ st
40
+ or 2
41
+ nd
42
+ degree relative, for example: 
43
+
44
+
45
+
46
+
47
+ Sudden death under 60 years of age
48
+
49
+
50
+ Heart transplant
51
+
52
+
53
+ Implantable cardiac defibrillator (ICD) under 60 years of age
54
+
55
+
56
+ Features of cardiomyopathy (e.g., systolic dysfunction, hypertrophy, left ventricular enlargement in an individual without risk factors).
57
+
58
+
59
+ Other related/overlapping cardiomyopathies
60
+
61
+
62
+
63
+
64
+ Examples of non-suspicious family history may include non-specific clinical features (e.g., palpitations, syncope, borderline/inconclusive echocardiogram findings, heart attack if age appropriate and suspected to result from coronary artery disease), but every attempt should be made to clarify features. 
65
+
66
+
67
+ Generally, this criterion is only applicable in the ABSENCE of any other possible disease-causing variants.  If other pathogenic or likely pathogenic variants are present, consider decreasing points assigned or overall weight.",Disease-specific
68
+ MYL3 (HGNC:7584),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
69
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
70
+ MYL3 (HGNC:7584),PS3,Strong,"In vitro splicing assays (e.g., RNA studies)
71
+
72
+
73
+ In vitro
74
+ splicing assays may be considered as 
75
+ STRONG
76
+ evidence, providing the following criteria are met.
77
+
78
+
79
+
80
+
81
+ Prior knowledge of predominant transcripts in cardiac tissue
82
+
83
+
84
+
85
+
86
+ Analysis undertaken using RNA extracted from cardiac tissue from the individual with the variant
87
+
88
+
89
+ Analysis undertaken using RNA extracted from whole blood providing the relevant transcripts (isoforms) are expressed in blood and are at sufficient levels to assess splice disruption.
90
+
91
+
92
+ Assay shows a clear, reproducible and convincing effect on splicing (i.e. a distinct splice product, present at a level comparable to the splice product from the wild-type allele), which is not observed in controls
93
+
94
+
95
+
96
+
97
+ Confirmation of abnormal splice product by Sanger sequencing
98
+
99
+
100
+
101
+
102
+ NOTE:
103
+ Mini-gene assay in non-patient derived cell lines are NOT considered to provide STRONG evidence.
104
+
105
+
106
+ NOTE:
107
+   Whether to activate this rule needs to be reconciled with the variant spectrum and disease mechanism for the gene at hand (i.e., consider whether the effect is likely to lead to LOF or an in-frame alteration and whether this type of effect is expected to be disease causing) (Abou Tayoun
108
+ et al.
109
+ 2018
110
+ 3
111
+ ).",Disease-specific
112
+ MYL3 (HGNC:7584),PS3,Moderate,"In vivo models (e.g., variant knock-in animal models)
113
+
114
+
115
+ Mammalian variant-specific knock-in animal models that produce a phenotype consistent with the clinical phenotype in humans (e.g., structural and/or functional cardiac abnormalities, premature death, arrhythmia) may be considered as
116
+ MODERATE
117
+ evidence
118
+
119
+
120
+ NOTE:
121
+ The following assays/models do NOT meet criteria
122
+
123
+
124
+
125
+
126
+ Assays that are known to be associated with non-specific cardiac phenotypes (e.g., morpholino-induced pericardial edema in zebrafish)
127
+
128
+
129
+ In vivo evidence that is not variant specific, such as whole gene alterations (i.e., cDNA or whole gene transgenic mice and whole or partial gene knock-out mice)",Disease-specific
130
+ MYL3 (HGNC:7584),PS3,Supporting,"In vitro
131
+
132
+ assays (e.g., biochemical assays of myofilament function, motility assays, human iPSC-CM)
133
+
134
+
135
+ While some
136
+ in vitro
137
+ assays may provide evidence that a variant in a cardiomyopathy gene has an effect on protein and/or myofilament function, at present, there are no validated “gold-standard” assays that are considered to reliably predict the clinical phenotype.
138
+
139
+
140
+ As such, in the cardiomyopathy genes listed in these guidelines, data from individual
141
+ in vitro
142
+ studies are unlikely to meet the criteria required to assign this rule at more than SUPPORTING level.",Disease-specific
143
+ MYL3 (HGNC:7584),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
144
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
145
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
146
+ MYL3 (HGNC:7584),PS4,Strong,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
147
+
148
+
149
+ Cohorts used in these analyses should meet the following criteria: 
150
+
151
+
152
+
153
+
154
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
155
+
156
+
157
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
158
+
159
+
160
+
161
+
162
+
163
+
164
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
165
+
166
+
167
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
168
+
169
+
170
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
171
+
172
+
173
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
174
+
175
+
176
+ The population diversity of the case and control cohorts are broadly similar.
177
+
178
+
179
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
180
+ et al.
181
+ 2017
182
+ 5
183
+ ,
184
+ DECIPHER
185
+ ).
186
+
187
+
188
+
189
+
190
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
191
+
192
+
193
+
194
+
195
+ STRONG
196
+ evidence requires the lower bound of the 95% confidence interval (CI) around the odds ratio (OR) estimate to be
197
+ ≥20
198
+
199
+
200
+
201
+
202
+ A PS4 calculator is available at
203
+ www.cardiodb.org
204
+ .
205
+
206
+
207
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
208
+
209
+
210
+ *RELEVANT PHENOTYPES:
211
+
212
+
213
+
214
+
215
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
216
+
217
+
218
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
219
+
220
+
221
+ Additional considerations for LVNC and end-stage HCM: 
222
+
223
+
224
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
225
+ et al.
226
+ 2017
227
+ 6
228
+ ; Oechslin
229
+ et al.
230
+ 2017
231
+ 7
232
+ ; Hershberger
233
+ et al.
234
+ 2017
235
+ 8
236
+ ; Ross
237
+ et al.
238
+ 2020
239
+ 9
240
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
241
+
242
+
243
+
244
+
245
+
246
+
247
+
248
+
249
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
250
+ MYL3 (HGNC:7584),PS4,Moderate,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
251
+
252
+
253
+ Cohorts used in these analyses should meet the following criteria: 
254
+
255
+
256
+
257
+
258
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
259
+
260
+
261
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
262
+
263
+
264
+
265
+
266
+
267
+
268
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
269
+
270
+
271
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
272
+
273
+
274
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
275
+
276
+
277
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
278
+
279
+
280
+ The population diversity of the case and control cohorts are broadly similar.
281
+
282
+
283
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
284
+ et al.
285
+ 2017
286
+ 5
287
+ ,
288
+ DECIPHER
289
+ ).
290
+
291
+
292
+
293
+
294
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
295
+
296
+
297
+
298
+
299
+ MODERATE
300
+ evidence requires the lower bound of the 95% CI around the OR to be
301
+ ≥10
302
+
303
+
304
+
305
+
306
+ A PS4 calculator is available at
307
+ www.cardiodb.org
308
+ .
309
+
310
+
311
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
312
+
313
+
314
+ *RELEVANT PHENOTYPES:
315
+
316
+
317
+
318
+
319
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
320
+
321
+
322
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
323
+
324
+
325
+ Additional considerations for LVNC and end-stage HCM: 
326
+
327
+
328
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
329
+ et al.
330
+ 2017
331
+ 6
332
+ ; Oechslin
333
+ et al.
334
+ 2017
335
+ 7
336
+ ; Hershberger
337
+ et al.
338
+ 2017
339
+ 8
340
+ ; Ross
341
+ et al.
342
+ 2020
343
+ 9
344
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
345
+
346
+
347
+
348
+
349
+
350
+
351
+
352
+
353
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
354
+ MYL3 (HGNC:7584),PS4,Supporting,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
355
+
356
+
357
+ Cohorts used in these analyses should meet the following criteria: 
358
+
359
+
360
+
361
+
362
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
363
+
364
+
365
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
366
+
367
+
368
+
369
+
370
+
371
+
372
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
373
+
374
+
375
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
376
+
377
+
378
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
379
+
380
+
381
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
382
+
383
+
384
+ The population diversity of the case and control cohorts are broadly similar.
385
+
386
+
387
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
388
+ et al.
389
+ 2017
390
+ 5
391
+ ,
392
+ DECIPHER
393
+ ).
394
+
395
+
396
+
397
+
398
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
399
+
400
+
401
+
402
+
403
+ SUPPORTING
404
+ evidence requires the lower bound of the 95% CI around the OR to be
405
+ ≥5
406
+
407
+
408
+
409
+
410
+ A PS4 calculator is available at
411
+ www.cardiodb.org
412
+ .
413
+
414
+
415
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
416
+
417
+
418
+ *RELEVANT PHENOTYPES:
419
+
420
+
421
+
422
+
423
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
424
+
425
+
426
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
427
+
428
+
429
+ Additional considerations for LVNC and end-stage HCM: 
430
+
431
+
432
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
433
+ et al.
434
+ 2017
435
+ 6
436
+ ; Oechslin
437
+ et al.
438
+ 2017
439
+ 7
440
+ ; Hershberger
441
+ et al.
442
+ 2017
443
+ 8
444
+ ; Ross
445
+ et al.
446
+ 2020
447
+ 9
448
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
449
+
450
+
451
+
452
+
453
+
454
+
455
+
456
+
457
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
458
+ MYL3 (HGNC:7584),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
459
+ MYL3 (HGNC:7584),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
460
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
461
+ MYL3 (HGNC:7584),PM2,Supporting,"The values used to calculate the PM2 thresholds were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/500 or lower), where the most frequent pathogenic variant accounts for no more than 2% of cases (e.g., has an allele frequency of ≤0.02 in cases based on the upper bound of 95% CI), and where the penetrance of a pathogenic variant is expected to be at least 50% (Kelly
462
+ et al.
463
+ 2018
464
+ 10
465
+ ).
466
+
467
+
468
+ A threshold of
469
+ ≤0.00004
470
+ in the subpopulation with the highest frequency when using the upper bound of the 95% CI activates this rule.
471
+
472
+
473
+
474
+
475
+ Alternatively, this is equivalent to the variant NOT being observed more than once (≤1 allele) in gnomAD v.2.1.1 in one of the non-founder populations (e.g., absence required from the Other and Ashkenazi Jewish subpopulations).
476
+
477
+
478
+ Applying a threshold of ≤0.00004 (upper bound of 95% CI of the allele frequency in gnomAD) is equivalent to the variant being seen in a single subpopulation and that subpopulation meets any of the following:
479
+
480
+
481
+ Allele Count (AC) in Allele Number (AN)
482
+
483
+
484
+ ≤1 in ≥120,000
485
+
486
+
487
+ ≤2 in ≥160,000
488
+
489
+
490
+ ≤3 in ≥195,000
491
+
492
+
493
+ ≤4 in ≥230,000
494
+
495
+
496
+
497
+
498
+
499
+
500
+
501
+
502
+ gnomAD is the preferred database for this calculation, but currently only displays the filtering allele frequency (FAF), which is equivalent to a lower bound estimate of the 95% CI, when the upper bound is what is needed.
503
+
504
+
505
+
506
+
507
+ Confidence interval tools, such as
508
+ Confit-de-MAF
509
+ , can be used to determine the upper bound of the 95% CI of the observed allele frequency.
510
+
511
+
512
+
513
+
514
+ Due to current technical limitations of next generation sequencing technologies, minor allele frequencies for complex variants (e.g., large indels) may not be accurately represented in population databases.
515
+
516
+
517
+ Caution should be used when a variant is only identified, or over-represented, in one of the smaller gnomAD populations, as the gnomAD allele frequencies may not accurately represent the true population frequency.
518
+
519
+
520
+ Population databases may contain affected or pre-symptomatic individuals for diseases with reduced penetrance/variable onset.",Disease-specific
521
+ MYL3 (HGNC:7584),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
522
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
523
+ MYL3 (HGNC:7584),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
524
+ MYL3 (HGNC:7584),PM4,Moderate,"Strength of rule should be carefully considered and may require downgrading to SUPPORTING based on the predicted impact of the variant, including the size of the deletion/insertion, its location, and conservation of the region. 
525
+
526
+
527
+ For genes where PVS1 is not applicable (i.e., where there is no evidence that pLOF variants cause disease), consider using this rule at MODERATE or SUPPORTING strength for truncating variants that do NOT undergo nonsense mediated decay (NMD).",General recommendation
528
+ MYL3 (HGNC:7584),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
529
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
530
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
531
+ MYL3 (HGNC:7584),PM5,Moderate,"This criterion can be used at MODERATE if a different missense variant at the same codon has been classified as
532
+ pathogenic
533
+ using these modified guidelines without application of PM5.
534
+
535
+
536
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to SUPPORTING if the predicted impact is not expected to be equivalent or more severe.
537
+
538
+
539
+ PM5 should not be combined with PM1.  If both are applicable at MODERATE weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
540
+ MYL3 (HGNC:7584),PM5,Supporting,"This criterion can be considered at SUPPORTING if a different missense variant at the same codon has been classified as
541
+ likely pathogenic
542
+ using these modified guidelines without application of PM5.
543
+
544
+
545
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as likely pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to NOT APPLICABLE if the predicted impact is not expected to be equivalent or more severe.
546
+
547
+
548
+ PM5 should not be combined with PM1.  The one with the higher strength should be applied, but if both are applicable at SUPPORTING weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
549
+ MYL3 (HGNC:7584),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
550
+ MYL3 (HGNC:7584),PM6,Moderate,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
551
+ 2
552
+ .
553
+
554
+
555
+ For most cardiomyopathies, it is recommended to default to “phenotype consistent with gene but not highly specific”. Clinical judgment is required for shifting to a higher or lower phenotypic consistency. 
556
+
557
+
558
+ See PS2 for additional considerations.",Disease-specific
559
+ MYL3 (HGNC:7584),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
560
+ Note: May be used as stronger evidence with increasing segregation data.",
561
+ MYL3 (HGNC:7584),PP1,Strong,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
562
+ ≥7
563
+
564
+ segregations
565
+ (LOD score of 2.1) for
566
+ STRONG
567
+ .
568
+
569
+
570
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
571
+ 11
572
+ can be considered: 
573
+
574
+
575
+
576
+
577
+ STRONG evidence requires ≥5 segregations (LOD score of 1.5)
578
+
579
+
580
+
581
+
582
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
583
+
584
+
585
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
586
+
587
+
588
+ Important considerations include:
589
+
590
+
591
+
592
+
593
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
594
+
595
+
596
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1.
597
+
598
+
599
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
600
+
601
+
602
+ Caution is needed when distantly related (≥3
603
+ rd
604
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
605
+ MYL3 (HGNC:7584),PP1,Moderate,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
606
+ ≥5
607
+
608
+ segregations
609
+ (LOD score of 1.5) for
610
+ MODERATE
611
+ .
612
+
613
+
614
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
615
+ 11
616
+ can be considered: 
617
+
618
+
619
+
620
+
621
+ MODERATE evidence requires ≥4 segregations (LOD score of 1.2)
622
+
623
+
624
+
625
+
626
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
627
+
628
+
629
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
630
+
631
+
632
+ Important considerations include:
633
+
634
+
635
+
636
+
637
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
638
+
639
+
640
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
641
+
642
+
643
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
644
+
645
+
646
+ Caution is needed when distantly related (≥3
647
+ rd
648
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
649
+ MYL3 (HGNC:7584),PP1,Supporting,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
650
+ ≥3
651
+
652
+ segregations
653
+ (LOD score of 0.9) for
654
+ SUPPORTING
655
+ . The thresholds as proposed by Jarvik and Browning (2016)
656
+ 11
657
+ are the same at ≥3 segregations (LOD score of 0.9) for supporting.
658
+
659
+
660
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
661
+
662
+
663
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
664
+
665
+
666
+ Important considerations include:
667
+
668
+
669
+
670
+
671
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
672
+
673
+
674
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
675
+
676
+
677
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
678
+
679
+
680
+ Caution is needed when distantly related (≥3
681
+ rd
682
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
683
+ MYL3 (HGNC:7584),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
684
+ MYL3 (HGNC:7584),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
685
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
686
+ MYL3 (HGNC:7584),PP3,Supporting,"As many
687
+ in silico
688
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
689
+
690
+
691
+ Use of REVEL (Ioannidis
692
+ et al.
693
+ 2016
694
+ 12
695
+ ) is recommended at thresholds of
696
+ ≥0.70 for PP3
697
+ .
698
+
699
+
700
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
701
+
702
+
703
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
704
+
705
+
706
+ SpliceAI
707
+ 13
708
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
709
+ MYL3 (HGNC:7584),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
710
+ MYL3 (HGNC:7584),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
711
+ MYL3 (HGNC:7584),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
712
+ MYL3 (HGNC:7584),BA1,Stand Alone,"Allele frequency is
713
+ ≥0.001
714
+ based on the
715
+ filtering allele frequency (FAF)
716
+ in
717
+ gnomAD
718
+ in the subpopulation with the highest frequency (popmax).
719
+
720
+
721
+ The values used to calculate the BA1 threshold were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/300 or lower).
722
+
723
+
724
+ The threshold is applicable when assessing variants in the context of autosomal dominant cardiomyopathy. 
725
+
726
+
727
+
728
+
729
+ Caution should be applied when assessing variants in the
730
+ MYL2
731
+ and
732
+ MYL3
733
+ genes, as homozygous or compound heterozygous variants have been reported to cause a recessive HCM and heterozygous individuals show no sign of disease.
734
+
735
+
736
+
737
+
738
+ gnomAD is the preferred database for this calculation. If a subpopulation specific FAF other than the popmax is needed, this value can be calculated using the AlleleFrequencyApp on the
739
+ CardioDB website
740
+ .
741
+
742
+
743
+
744
+
745
+ Using the Inverse AF tab, enter in the population size and the number of alleles identified and it will calculate the FAF.  
746
+
747
+
748
+ Set confidence to 0.95 (95%).
749
+
750
+
751
+ If the FAF is ≥0.001, this rule can be applied.
752
+
753
+
754
+
755
+
756
+ The FAF by platform (e.g., exome vs. genome; v.2.1.1 vs. v.3.1.1) should be considered, the larger population is most likely to have the most accurate representation of “true” population allele frequency.
757
+
758
+
759
+ Caution is needed when considering any population cohorts that are smaller than the smallest subpopulations within gnomAD v.2.1.1 (e.g., ~5000 individuals or ~10,000 alleles). Despite this conservative nature of this threshold and approach, in smaller cohorts, the observed allele frequency may less accurately reflect the true allele frequency. Traditionally, once a variant is classified as Benign, it is rarely re-evaluated and so the highest confidence is needed to establish that classification on an allele frequency alone.",Disease-specific
760
+ MYL3 (HGNC:7584),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
761
+ MYL3 (HGNC:7584),BS1,Strong,"Allele frequency is
762
+ ≥0.0001 for
763
+
764
+ MYL3
765
+ based on the
766
+ filtering allele frequency (FAF)
767
+ in
768
+ gnomAD
769
+ in the subpopulation with the highest frequency (popmax).
770
+
771
+
772
+ Criterion BS1 may only be used as standalone evidence to classify a variant as Likely Benign in the absence of conflicting data. See SVI guidance (Tavtigian
773
+ et al.
774
+ 2018
775
+ 14
776
+ ; Tavtigian
777
+ et al.
778
+ 2020
779
+ 15
780
+ ). 
781
+
782
+
783
+ See BA1 for additional specifications that also apply to BS1.","Disease-specific,Gene-specific"
784
+ MYL3 (HGNC:7584),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",NA
785
+ MYL3 (HGNC:7584),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
786
+ MYL3 (HGNC:7584),BS3,Strong,See PS3 specifications.,Disease-specific
787
+ MYL3 (HGNC:7584),BS3,Moderate,See PS3 specifications.,Disease-specific
788
+ MYL3 (HGNC:7584),BS3,Supporting,See PS3 specifications.,Disease-specific
789
+ MYL3 (HGNC:7584),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
790
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
791
+ MYL3 (HGNC:7584),BS4,Strong,"Any non-segregations should be carefully evaluated to rule out a phenocopy or the presence of a second disease-causing variant before considering it as conflicting or benign evidence. 
792
+
793
+
794
+
795
+
796
+ The presence of “phenocopies” (e.g., athlete’s heart, hypertensive heart disease, ischemic cardiomyopathy, alcoholic cardiomyopathy, diabetic cardiomyopathy) can mimic non-segregation (i.e., lack of segregation) among affected individuals. 
797
+
798
+
799
+ Families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent ‘non-segregation’.
800
+
801
+
802
+
803
+
804
+ Because of these possibilities,
805
+ multiple (≥2) non-segregations
806
+ that are highly unlikely to be phenocopies or due to alternate variants (e.g., those without a possible alternate cause)
807
+ are required to apply this rule
808
+ .  A higher number of non-segregations is necessary for instances where alternative causes are possible (e.g., non-segregation in a sibling with childhood onset cardiomyopathy versus a grandparent with hypertension and HCM).
809
+
810
+
811
+ Careful consideration of the above points is required when using this data as conflicting evidence, especially when overall evidence supports likely pathogenic or pathogenic.",Disease-specific
812
+ MYL3 (HGNC:7584),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
813
+ MYL3 (HGNC:7584),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
814
+ MYL3 (HGNC:7584),BP2,Supporting,"Other variants must be pathogenic as defined by these specifications.
815
+
816
+
817
+ Testing of parents or other informative relatives is often required to determine
818
+ cis
819
+ /
820
+ trans
821
+ status.
822
+
823
+
824
+ If a variant is seen in
825
+ trans
826
+ (or as double heterozygous) with another pathogenic variant in ≥2 cases and the phenotype is not more severe than when either of the two variants are seen in isolation, this rule may be applied (i.e., high confidence this variant is NOT contributing to disease).
827
+
828
+
829
+
830
+
831
+ <1% of cases of HCM have >1 pathogenic or likely pathogenic variant (0.6%; Alfares
832
+ et al.
833
+ 2015
834
+ 16
835
+ ).
836
+
837
+
838
+
839
+
840
+ This rule cannot be applied when the variant has only been observed in
841
+ cis
842
+ with a pathogenic variant as its significance in isolation is unknown in this scenario. 
843
+
844
+
845
+ Caution is needed if using this criterion as a primary piece of evidence for classifying a variant as likely benign/benign (i.e., only 2 SUPPORTING criteria are sufficient for a likely benign classification).",Disease-specific
846
+ MYL3 (HGNC:7584),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
847
+ MYL3 (HGNC:7584),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
848
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
849
+ MYL3 (HGNC:7584),BP4,Supporting,"As many
850
+ in silico
851
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
852
+
853
+
854
+ Use of REVEL (Ioannidis et al. 2016
855
+ 12
856
+ ) is recommended at thresholds of
857
+ ≤0.40 for BP4
858
+ .
859
+
860
+
861
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
862
+
863
+
864
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
865
+
866
+
867
+ SpliceAI
868
+ 13
869
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
870
+ MYL3 (HGNC:7584),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
871
+ MYL3 (HGNC:7584),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
872
+ MYL3 (HGNC:7584),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
873
+ MYL3 (HGNC:7584),BP7,Supporting,"Also applicable to
874
+ intronic variants outside the splice consensus sequence (-4 and +7 outward)
875
+ for which splicing prediction algorithms predict no impact to the splice consensus sequence NOR the creation of a new splice site AND the nucleotide is not highly conserved.
876
+
877
+
878
+ Rule can be combined with BP4 to make a variant likely benign per Richards
879
+ et al.
880
+ 2015
881
+ 1
882
+ .",General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTNNI3Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,897 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ TNNI3 (HGNC:11947),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",NA
8
+ TNNI3 (HGNC:11947),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
9
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
10
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
11
+ TNNI3 (HGNC:11947),PS1,Strong,"No cardiomyopathy specifications. Apply as outlined by Richards
12
+ et al
13
+ . 2015
14
+ 1
15
+ .
16
+
17
+
18
+ Example of when rule should NOT be applied. NM_000256.3(
19
+ MYBPC3
20
+ ): c.2308G>A (p.Asp770Asn) has an established impact on splicing leading to nonsense mediated decay (NMD) and should not be used to provide evidence for other variants observed to result in the same amino acid change.",No change
21
+ TNNI3 (HGNC:11947),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
22
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
23
+ TNNI3 (HGNC:11947),PS2,Strong,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
24
+ 2
25
+ .
26
+
27
+
28
+ For most cardiomyopathies, it is recommended to default to
29
+ Phenotype consistency: “Phenotype consistent with gene but not highly specific”
30
+ . Clinical judgment is required for shifting to a higher or lower category. 
31
+
32
+
33
+ For use as a STRONG or VERY STRONG criterion, ideally parents have been thoroughly clinically evaluated without evidence of cardiomyopathy (ideally using a combination of ECG and echocardiogram or cardiac MRI for maximum sensitivity).
34
+
35
+
36
+ A family history consistent with
37
+ de novo
38
+ inheritance should not have any clinical signs or symptoms suggestive of cardiomyopathy in a 1
39
+ st
40
+ or 2
41
+ nd
42
+ degree relative, for example: 
43
+
44
+
45
+
46
+
47
+ Sudden death under 60 years of age
48
+
49
+
50
+ Heart transplant
51
+
52
+
53
+ Implantable cardiac defibrillator (ICD) under 60 years of age
54
+
55
+
56
+ Features of cardiomyopathy (e.g., systolic dysfunction, hypertrophy, left ventricular enlargement in an individual without risk factors).
57
+
58
+
59
+ Other related/overlapping cardiomyopathies
60
+
61
+
62
+
63
+
64
+ Examples of non-suspicious family history may include non-specific clinical features (e.g., palpitations, syncope, borderline/inconclusive echocardiogram findings, heart attack if age appropriate and suspected to result from coronary artery disease), but every attempt should be made to clarify features. 
65
+
66
+
67
+ Generally, this criterion is only applicable in the ABSENCE of any other possible disease-causing variants.  If other pathogenic or likely pathogenic variants are present, consider decreasing points assigned or overall weight.",Disease-specific
68
+ TNNI3 (HGNC:11947),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
69
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
70
+ TNNI3 (HGNC:11947),PS3,Strong,"In vitro splicing assays (e.g., RNA studies)
71
+
72
+
73
+ In vitro
74
+ splicing assays may be considered as 
75
+ STRONG
76
+ evidence, providing the following criteria are met.
77
+
78
+
79
+
80
+
81
+ Prior knowledge of predominant transcripts in cardiac tissue
82
+
83
+
84
+
85
+
86
+ Analysis undertaken using RNA extracted from cardiac tissue from the individual with the variant
87
+
88
+
89
+ Analysis undertaken using RNA extracted from whole blood providing the relevant transcripts (isoforms) are expressed in blood and are at sufficient levels to assess splice disruption.
90
+
91
+
92
+ Assay shows a clear, reproducible and convincing effect on splicing (i.e. a distinct splice product, present at a level comparable to the splice product from the wild-type allele), which is not observed in controls
93
+
94
+
95
+
96
+
97
+ Confirmation of abnormal splice product by Sanger sequencing
98
+
99
+
100
+
101
+
102
+ NOTE:
103
+ Mini-gene assay in non-patient derived cell lines are NOT considered to provide STRONG evidence.
104
+
105
+
106
+ NOTE:
107
+   Whether to activate this rule needs to be reconciled with the variant spectrum and disease mechanism for the gene at hand (i.e., consider whether the effect is likely to lead to LOF or an in-frame alteration and whether this type of effect is expected to be disease causing) (Abou Tayoun
108
+ et al.
109
+ 2018
110
+ 3
111
+ ).",Disease-specific
112
+ TNNI3 (HGNC:11947),PS3,Moderate,"In vivo models (e.g., variant knock-in animal models)
113
+
114
+
115
+ Mammalian variant-specific knock-in animal models that produce a phenotype consistent with the clinical phenotype in humans (e.g., structural and/or functional cardiac abnormalities, premature death, arrhythmia) may be considered as
116
+ MODERATE
117
+ evidence
118
+
119
+
120
+ NOTE:
121
+ The following assays/models do NOT meet criteria
122
+
123
+
124
+
125
+
126
+ Assays that are known to be associated with non-specific cardiac phenotypes (e.g., morpholino-induced pericardial edema in zebrafish)
127
+
128
+
129
+ In vivo evidence that is not variant specific, such as whole gene alterations (i.e., cDNA or whole gene transgenic mice and whole or partial gene knock-out mice)",Disease-specific
130
+ TNNI3 (HGNC:11947),PS3,Supporting,"In vitro
131
+
132
+ assays (e.g., biochemical assays of myofilament function, motility assays, human iPSC-CM)
133
+
134
+
135
+ While some
136
+ in vitro
137
+ assays may provide evidence that a variant in a cardiomyopathy gene has an effect on protein and/or myofilament function, at present, there are no validated “gold-standard” assays that are considered to reliably predict the clinical phenotype.
138
+
139
+
140
+ As such, in the cardiomyopathy genes listed in these guidelines, data from individual
141
+ in vitro
142
+ studies are unlikely to meet the criteria required to assign this rule at more than SUPPORTING level.",Disease-specific
143
+ TNNI3 (HGNC:11947),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
144
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
145
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
146
+ TNNI3 (HGNC:11947),PS4,Strong,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
147
+
148
+
149
+ Cohorts used in these analyses should meet the following criteria: 
150
+
151
+
152
+
153
+
154
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
155
+
156
+
157
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
158
+
159
+
160
+
161
+
162
+
163
+
164
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
165
+
166
+
167
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
168
+
169
+
170
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
171
+
172
+
173
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
174
+
175
+
176
+ The population diversity of the case and control cohorts are broadly similar.
177
+
178
+
179
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
180
+ et al.
181
+ 2017
182
+ 5
183
+ ,
184
+ DECIPHER
185
+ ).
186
+
187
+
188
+
189
+
190
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
191
+
192
+
193
+
194
+
195
+ STRONG
196
+ evidence requires the lower bound of the 95% confidence interval (CI) around the odds ratio (OR) estimate to be
197
+ ≥20
198
+
199
+
200
+
201
+
202
+ A PS4 calculator is available at
203
+ www.cardiodb.org
204
+ .
205
+
206
+
207
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
208
+
209
+
210
+ *RELEVANT PHENOTYPES:
211
+
212
+
213
+
214
+
215
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
216
+
217
+
218
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
219
+
220
+
221
+ Additional considerations for LVNC and end-stage HCM: 
222
+
223
+
224
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
225
+ et al.
226
+ 2017
227
+ 6
228
+ ; Oechslin
229
+ et al.
230
+ 2017
231
+ 7
232
+ ; Hershberger
233
+ et al.
234
+ 2017
235
+ 8
236
+ ; Ross
237
+ et al.
238
+ 2020
239
+ 9
240
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
241
+
242
+
243
+
244
+
245
+
246
+
247
+
248
+
249
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
250
+ TNNI3 (HGNC:11947),PS4,Moderate,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
251
+
252
+
253
+ Cohorts used in these analyses should meet the following criteria: 
254
+
255
+
256
+
257
+
258
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
259
+
260
+
261
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
262
+
263
+
264
+
265
+
266
+
267
+
268
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
269
+
270
+
271
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
272
+
273
+
274
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
275
+
276
+
277
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
278
+
279
+
280
+ The population diversity of the case and control cohorts are broadly similar.
281
+
282
+
283
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
284
+ et al.
285
+ 2017
286
+ 5
287
+ ,
288
+ DECIPHER
289
+ ).
290
+
291
+
292
+
293
+
294
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
295
+
296
+
297
+
298
+
299
+ MODERATE
300
+ evidence requires the lower bound of the 95% CI around the OR to be
301
+ ≥10
302
+
303
+
304
+
305
+
306
+ A PS4 calculator is available at
307
+ www.cardiodb.org
308
+ .
309
+
310
+
311
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
312
+
313
+
314
+ *RELEVANT PHENOTYPES:
315
+
316
+
317
+
318
+
319
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
320
+
321
+
322
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
323
+
324
+
325
+ Additional considerations for LVNC and end-stage HCM: 
326
+
327
+
328
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
329
+ et al.
330
+ 2017
331
+ 6
332
+ ; Oechslin
333
+ et al.
334
+ 2017
335
+ 7
336
+ ; Hershberger
337
+ et al.
338
+ 2017
339
+ 8
340
+ ; Ross
341
+ et al.
342
+ 2020
343
+ 9
344
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
345
+
346
+
347
+
348
+
349
+
350
+
351
+
352
+
353
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
354
+ TNNI3 (HGNC:11947),PS4,Supporting,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
355
+
356
+
357
+ Cohorts used in these analyses should meet the following criteria: 
358
+
359
+
360
+
361
+
362
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
363
+
364
+
365
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
366
+
367
+
368
+
369
+
370
+
371
+
372
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
373
+
374
+
375
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
376
+
377
+
378
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
379
+
380
+
381
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
382
+
383
+
384
+ The population diversity of the case and control cohorts are broadly similar.
385
+
386
+
387
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
388
+ et al.
389
+ 2017
390
+ 5
391
+ ,
392
+ DECIPHER
393
+ ).
394
+
395
+
396
+
397
+
398
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
399
+
400
+
401
+
402
+
403
+ SUPPORTING
404
+ evidence requires the lower bound of the 95% CI around the OR to be
405
+ ≥5
406
+
407
+
408
+
409
+
410
+ A PS4 calculator is available at
411
+ www.cardiodb.org
412
+ .
413
+
414
+
415
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
416
+
417
+
418
+ *RELEVANT PHENOTYPES:
419
+
420
+
421
+
422
+
423
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
424
+
425
+
426
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
427
+
428
+
429
+ Additional considerations for LVNC and end-stage HCM: 
430
+
431
+
432
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
433
+ et al.
434
+ 2017
435
+ 6
436
+ ; Oechslin
437
+ et al.
438
+ 2017
439
+ 7
440
+ ; Hershberger
441
+ et al.
442
+ 2017
443
+ 8
444
+ ; Ross
445
+ et al.
446
+ 2020
447
+ 9
448
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
449
+
450
+
451
+
452
+
453
+
454
+
455
+
456
+
457
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
458
+ TNNI3 (HGNC:11947),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
459
+ TNNI3 (HGNC:11947),PM1,Moderate,"Applicable to missense variants in
460
+ TNNI3
461
+ in the specific regions listed below (Walsh 
462
+ et al.
463
+ 2019
464
+ 10
465
+ ). 
466
+
467
+
468
+
469
+
470
+ Transcripts ENST00000344887 and NM_000363.5
471
+
472
+
473
+ Codons 141-209
474
+
475
+
476
+
477
+
478
+ Data from HCM case cohorts was used to derive these cluster regions. Therefore, this rule should NOT be applied when additional evidence for the variant supports that the variant causes a phenotype other than HCM (e.g., variant seen in multiple DCM cases).
479
+
480
+
481
+ Enrichment was not observed for DCM in any genes.
482
+
483
+
484
+ Rule should NOT be combined with PM5 because presence of pathogenic variants in the same codon/region were used to determine clustering and would be double-counting evidence.","Disease-specific,Gene-specific"
485
+ TNNI3 (HGNC:11947),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
486
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
487
+ TNNI3 (HGNC:11947),PM2,Supporting,"The values used to calculate the PM2 thresholds were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/500 or lower), where the most frequent pathogenic variant accounts for no more than 2% of cases (e.g., has an allele frequency of ≤0.02 in cases based on the upper bound of 95% CI), and where the penetrance of a pathogenic variant is expected to be at least 50% (Kelly
488
+ et al.
489
+ 2018
490
+ 11
491
+ ).
492
+
493
+
494
+ A threshold of
495
+ ≤0.00004
496
+ in the subpopulation with the highest frequency when using the upper bound of the 95% CI activates this rule.
497
+
498
+
499
+
500
+
501
+ Alternatively, this is equivalent to the variant NOT being observed more than once (≤1 allele) in gnomAD v.2.1.1 in one of the non-founder populations (e.g., absence required from the Other and Ashkenazi Jewish subpopulations).
502
+
503
+
504
+ Applying a threshold of ≤0.00004 (upper bound of 95% CI of the allele frequency in gnomAD) is equivalent to the variant being seen in a single subpopulation and that subpopulation meets any of the following:
505
+
506
+
507
+ Allele Count (AC) in Allele Number (AN)
508
+
509
+
510
+ ≤1 in ≥120,000
511
+
512
+
513
+ ≤2 in ≥160,000
514
+
515
+
516
+ ≤3 in ≥195,000
517
+
518
+
519
+ ≤4 in ≥230,000
520
+
521
+
522
+
523
+
524
+
525
+
526
+
527
+
528
+ gnomAD is the preferred database for this calculation, but currently only displays the filtering allele frequency (FAF), which is equivalent to a lower bound estimate of the 95% CI, when the upper bound is what is needed.
529
+
530
+
531
+
532
+
533
+ Confidence interval tools, such as
534
+ Confit-de-MAF
535
+ , can be used to determine the upper bound of the 95% CI of the observed allele frequency.
536
+
537
+
538
+
539
+
540
+ Due to current technical limitations of next generation sequencing technologies, minor allele frequencies for complex variants (e.g., large indels) may not be accurately represented in population databases.
541
+
542
+
543
+ Caution should be used when a variant is only identified, or over-represented, in one of the smaller gnomAD populations, as the gnomAD allele frequencies may not accurately represent the true population frequency.
544
+
545
+
546
+ Population databases may contain affected or pre-symptomatic individuals for diseases with reduced penetrance/variable onset.",Disease-specific
547
+ TNNI3 (HGNC:11947),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
548
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
549
+ TNNI3 (HGNC:11947),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
550
+ TNNI3 (HGNC:11947),PM4,Moderate,"Strength of rule should be carefully considered and may require downgrading to SUPPORTING based on the predicted impact of the variant, including the size of the deletion/insertion, its location, and conservation of the region. 
551
+
552
+
553
+ For genes where PVS1 is not applicable (i.e., where there is no evidence that pLOF variants cause disease), consider using this rule at MODERATE or SUPPORTING strength for truncating variants that do NOT undergo nonsense mediated decay (NMD).",General recommendation
554
+ TNNI3 (HGNC:11947),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
555
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
556
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
557
+ TNNI3 (HGNC:11947),PM5,Moderate,"This criterion can be used at MODERATE if a different missense variant at the same codon has been classified as
558
+ pathogenic
559
+ using these modified guidelines without application of PM5.
560
+
561
+
562
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to SUPPORTING if the predicted impact is not expected to be equivalent or more severe.
563
+
564
+
565
+ PM5 should not be combined with PM1.  If both are applicable at MODERATE weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
566
+ TNNI3 (HGNC:11947),PM5,Supporting,"This criterion can be considered at SUPPORTING if a different missense variant at the same codon has been classified as
567
+ likely pathogenic
568
+ using these modified guidelines without application of PM5.
569
+
570
+
571
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as likely pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to NOT APPLICABLE if the predicted impact is not expected to be equivalent or more severe.
572
+
573
+
574
+ PM5 should not be combined with PM1.  The one with the higher strength should be applied, but if both are applicable at SUPPORTING weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
575
+ TNNI3 (HGNC:11947),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
576
+ TNNI3 (HGNC:11947),PM6,Moderate,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
577
+ 2
578
+ .
579
+
580
+
581
+ For most cardiomyopathies, it is recommended to default to “phenotype consistent with gene but not highly specific”. Clinical judgment is required for shifting to a higher or lower phenotypic consistency. 
582
+
583
+
584
+ See PS2 for additional considerations.",Disease-specific
585
+ TNNI3 (HGNC:11947),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
586
+ Note: May be used as stronger evidence with increasing segregation data.",
587
+ TNNI3 (HGNC:11947),PP1,Strong,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
588
+ ≥7
589
+
590
+ segregations
591
+ (LOD score of 2.1) for
592
+ STRONG
593
+ .
594
+
595
+
596
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
597
+ 12
598
+ can be considered: 
599
+
600
+
601
+
602
+
603
+ STRONG evidence requires ≥5 segregations (LOD score of 1.5)
604
+
605
+
606
+
607
+
608
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
609
+
610
+
611
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
612
+
613
+
614
+ Important considerations include:
615
+
616
+
617
+
618
+
619
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
620
+
621
+
622
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1.
623
+
624
+
625
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
626
+
627
+
628
+ Caution is needed when distantly related (≥3
629
+ rd
630
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
631
+ TNNI3 (HGNC:11947),PP1,Moderate,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
632
+ ≥5
633
+
634
+ segregations
635
+ (LOD score of 1.5) for
636
+ MODERATE
637
+ .
638
+
639
+
640
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
641
+ 12
642
+ can be considered: 
643
+
644
+
645
+
646
+
647
+ MODERATE evidence requires ≥4 segregations (LOD score of 1.2)
648
+
649
+
650
+
651
+
652
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
653
+
654
+
655
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
656
+
657
+
658
+ Important considerations include:
659
+
660
+
661
+
662
+
663
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
664
+
665
+
666
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
667
+
668
+
669
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
670
+
671
+
672
+ Caution is needed when distantly related (≥3
673
+ rd
674
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
675
+ TNNI3 (HGNC:11947),PP1,Supporting,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
676
+ ≥3
677
+
678
+ segregations
679
+ (LOD score of 0.9) for
680
+ SUPPORTING
681
+ . The thresholds as proposed by Jarvik and Browning (2016)
682
+ 12
683
+ are the same at ≥3 segregations (LOD score of 0.9) for supporting.
684
+
685
+
686
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
687
+
688
+
689
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
690
+
691
+
692
+ Important considerations include:
693
+
694
+
695
+
696
+
697
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
698
+
699
+
700
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
701
+
702
+
703
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
704
+
705
+
706
+ Caution is needed when distantly related (≥3
707
+ rd
708
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
709
+ TNNI3 (HGNC:11947),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
710
+ TNNI3 (HGNC:11947),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
711
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
712
+ TNNI3 (HGNC:11947),PP3,Supporting,"As many
713
+ in silico
714
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
715
+
716
+
717
+ Use of REVEL (Ioannidis
718
+ et al.
719
+ 2016
720
+ 13
721
+ ) is recommended at thresholds of
722
+ ≥0.70 for PP3
723
+ .
724
+
725
+
726
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
727
+
728
+
729
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
730
+
731
+
732
+ SpliceAI
733
+ 14
734
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
735
+ TNNI3 (HGNC:11947),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
736
+ TNNI3 (HGNC:11947),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
737
+ TNNI3 (HGNC:11947),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
738
+ TNNI3 (HGNC:11947),BA1,Stand Alone,"Allele frequency is
739
+ ≥0.001
740
+ based on the
741
+ filtering allele frequency (FAF)
742
+ in
743
+ gnomAD
744
+ in the subpopulation with the highest frequency (popmax).
745
+
746
+
747
+ The values used to calculate the BA1 threshold were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/300 or lower).
748
+
749
+
750
+ The threshold is applicable when assessing variants in the context of autosomal dominant cardiomyopathy. 
751
+
752
+
753
+ gnomAD is the preferred database for this calculation. If a subpopulation specific FAF other than the popmax is needed, this value can be calculated using the AlleleFrequencyApp on the
754
+ CardioDB website
755
+ .
756
+
757
+
758
+
759
+
760
+ Using the Inverse AF tab, enter in the population size and the number of alleles identified and it will calculate the FAF.  
761
+
762
+
763
+ Set confidence to 0.95 (95%).
764
+
765
+
766
+ If the FAF is ≥0.001, this rule can be applied.
767
+
768
+
769
+
770
+
771
+ The FAF by platform (e.g., exome vs. genome; v.2.1.1 vs. v.3.1.1) should be considered, the larger population is most likely to have the most accurate representation of “true” population allele frequency.
772
+
773
+
774
+ Caution is needed when considering any population cohorts that are smaller than the smallest subpopulations within gnomAD v.2.1.1 (e.g., ~5000 individuals or ~10,000 alleles). Despite this conservative nature of this threshold and approach, in smaller cohorts, the observed allele frequency may less accurately reflect the true allele frequency. Traditionally, once a variant is classified as Benign, it is rarely re-evaluated and so the highest confidence is needed to establish that classification on an allele frequency alone.",Disease-specific
775
+ TNNI3 (HGNC:11947),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
776
+ TNNI3 (HGNC:11947),BS1,Strong,"Allele frequency is
777
+ ≥0.0001 for
778
+
779
+ TNNI3
780
+ based on the
781
+ filtering allele frequency (FAF)
782
+ in
783
+ gnomAD
784
+ in the subpopulation with the highest frequency (popmax).
785
+
786
+
787
+ Criterion BS1 may only be used as standalone evidence to classify a variant as Likely Benign in the absence of conflicting data. See SVI guidance (Tavtigian
788
+ et al.
789
+ 2018
790
+ 15
791
+ ; Tavtigian
792
+ et al.
793
+ 2020
794
+ 16
795
+ ). 
796
+
797
+
798
+ See BA1 for additional specifications that also apply to BS1.","Disease-specific,Gene-specific"
799
+ TNNI3 (HGNC:11947),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",NA
800
+ TNNI3 (HGNC:11947),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
801
+ TNNI3 (HGNC:11947),BS3,Strong,See PS3 specifications.,Disease-specific
802
+ TNNI3 (HGNC:11947),BS3,Moderate,See PS3 specifications.,Disease-specific
803
+ TNNI3 (HGNC:11947),BS3,Supporting,See PS3 specifications.,Disease-specific
804
+ TNNI3 (HGNC:11947),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
805
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
806
+ TNNI3 (HGNC:11947),BS4,Strong,"Any non-segregations should be carefully evaluated to rule out a phenocopy or the presence of a second disease-causing variant before considering it as conflicting or benign evidence. 
807
+
808
+
809
+
810
+
811
+ The presence of “phenocopies” (e.g., athlete’s heart, hypertensive heart disease, ischemic cardiomyopathy, alcoholic cardiomyopathy, diabetic cardiomyopathy) can mimic non-segregation (i.e., lack of segregation) among affected individuals. 
812
+
813
+
814
+ Families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent ‘non-segregation’.
815
+
816
+
817
+
818
+
819
+ Because of these possibilities,
820
+ multiple (≥2) non-segregations
821
+ that are highly unlikely to be phenocopies or due to alternate variants (e.g., those without a possible alternate cause)
822
+ are required to apply this rule
823
+ .  A higher number of non-segregations is necessary for instances where alternative causes are possible (e.g., non-segregation in a sibling with childhood onset cardiomyopathy versus a grandparent with hypertension and HCM).
824
+
825
+
826
+ Careful consideration of the above points is required when using this data as conflicting evidence, especially when overall evidence supports likely pathogenic or pathogenic.",Disease-specific
827
+ TNNI3 (HGNC:11947),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
828
+ TNNI3 (HGNC:11947),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
829
+ TNNI3 (HGNC:11947),BP2,Supporting,"Other variants must be pathogenic as defined by these specifications.
830
+
831
+
832
+ Testing of parents or other informative relatives is often required to determine
833
+ cis
834
+ /
835
+ trans
836
+ status.
837
+
838
+
839
+ If a variant is seen in
840
+ trans
841
+ (or as double heterozygous) with another pathogenic variant in ≥2 cases and the phenotype is not more severe than when either of the two variants are seen in isolation, this rule may be applied (i.e., high confidence this variant is NOT contributing to disease).
842
+
843
+
844
+
845
+
846
+ <1% of cases of HCM have >1 pathogenic or likely pathogenic variant (0.6%; Alfares
847
+ et al.
848
+ 2015
849
+ 17
850
+ ).
851
+
852
+
853
+
854
+
855
+ This rule cannot be applied when the variant has only been observed in
856
+ cis
857
+ with a pathogenic variant as its significance in isolation is unknown in this scenario. 
858
+
859
+
860
+ Caution is needed if using this criterion as a primary piece of evidence for classifying a variant as likely benign/benign (i.e., only 2 SUPPORTING criteria are sufficient for a likely benign classification).",Disease-specific
861
+ TNNI3 (HGNC:11947),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
862
+ TNNI3 (HGNC:11947),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
863
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
864
+ TNNI3 (HGNC:11947),BP4,Supporting,"As many
865
+ in silico
866
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
867
+
868
+
869
+ Use of REVEL (Ioannidis et al. 2016
870
+ 13
871
+ ) is recommended at thresholds of
872
+ ≤0.40 for BP4
873
+ .
874
+
875
+
876
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
877
+
878
+
879
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
880
+
881
+
882
+ SpliceAI
883
+ 14
884
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
885
+ TNNI3 (HGNC:11947),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
886
+ TNNI3 (HGNC:11947),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
887
+ TNNI3 (HGNC:11947),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
888
+ TNNI3 (HGNC:11947),BP7,Supporting,"Also applicable to
889
+ intronic variants outside the splice consensus sequence (-4 and +7 outward)
890
+ for which splicing prediction algorithms predict no impact to the splice consensus sequence NOR the creation of a new splice site AND the nucleotide is not highly conserved.
891
+
892
+
893
+ Rule can be combined with BP4 to make a variant likely benign per Richards
894
+ et al.
895
+ 2015
896
+ 1
897
+ .",General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTNNT2Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,897 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ TNNT2 (HGNC:11949),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",NA
8
+ TNNT2 (HGNC:11949),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
9
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
10
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
11
+ TNNT2 (HGNC:11949),PS1,Strong,"No cardiomyopathy specifications. Apply as outlined by Richards
12
+ et al
13
+ . 2015
14
+ 1
15
+ .
16
+
17
+
18
+ Example of when rule should NOT be applied. NM_000256.3(
19
+ MYBPC3
20
+ ): c.2308G>A (p.Asp770Asn) has an established impact on splicing leading to nonsense mediated decay (NMD) and should not be used to provide evidence for other variants observed to result in the same amino acid change.",No change
21
+ TNNT2 (HGNC:11949),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
22
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
23
+ TNNT2 (HGNC:11949),PS2,Strong,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
24
+ 2
25
+ .
26
+
27
+
28
+ For most cardiomyopathies, it is recommended to default to
29
+ Phenotype consistency: “Phenotype consistent with gene but not highly specific”
30
+ . Clinical judgment is required for shifting to a higher or lower category. 
31
+
32
+
33
+ For use as a STRONG or VERY STRONG criterion, ideally parents have been thoroughly clinically evaluated without evidence of cardiomyopathy (ideally using a combination of ECG and echocardiogram or cardiac MRI for maximum sensitivity).
34
+
35
+
36
+ A family history consistent with
37
+ de novo
38
+ inheritance should not have any clinical signs or symptoms suggestive of cardiomyopathy in a 1
39
+ st
40
+ or 2
41
+ nd
42
+ degree relative, for example: 
43
+
44
+
45
+
46
+
47
+ Sudden death under 60 years of age
48
+
49
+
50
+ Heart transplant
51
+
52
+
53
+ Implantable cardiac defibrillator (ICD) under 60 years of age
54
+
55
+
56
+ Features of cardiomyopathy (e.g., systolic dysfunction, hypertrophy, left ventricular enlargement in an individual without risk factors).
57
+
58
+
59
+ Other related/overlapping cardiomyopathies
60
+
61
+
62
+
63
+
64
+ Examples of non-suspicious family history may include non-specific clinical features (e.g., palpitations, syncope, borderline/inconclusive echocardiogram findings, heart attack if age appropriate and suspected to result from coronary artery disease), but every attempt should be made to clarify features. 
65
+
66
+
67
+ Generally, this criterion is only applicable in the ABSENCE of any other possible disease-causing variants.  If other pathogenic or likely pathogenic variants are present, consider decreasing points assigned or overall weight.",Disease-specific
68
+ TNNT2 (HGNC:11949),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
69
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
70
+ TNNT2 (HGNC:11949),PS3,Strong,"In vitro splicing assays (e.g., RNA studies)
71
+
72
+
73
+ In vitro
74
+ splicing assays may be considered as 
75
+ STRONG
76
+ evidence, providing the following criteria are met.
77
+
78
+
79
+
80
+
81
+ Prior knowledge of predominant transcripts in cardiac tissue
82
+
83
+
84
+
85
+
86
+ Analysis undertaken using RNA extracted from cardiac tissue from the individual with the variant
87
+
88
+
89
+ Analysis undertaken using RNA extracted from whole blood providing the relevant transcripts (isoforms) are expressed in blood and are at sufficient levels to assess splice disruption.
90
+
91
+
92
+ Assay shows a clear, reproducible and convincing effect on splicing (i.e. a distinct splice product, present at a level comparable to the splice product from the wild-type allele), which is not observed in controls
93
+
94
+
95
+
96
+
97
+ Confirmation of abnormal splice product by Sanger sequencing
98
+
99
+
100
+
101
+
102
+ NOTE:
103
+ Mini-gene assay in non-patient derived cell lines are NOT considered to provide STRONG evidence.
104
+
105
+
106
+ NOTE:
107
+   Whether to activate this rule needs to be reconciled with the variant spectrum and disease mechanism for the gene at hand (i.e., consider whether the effect is likely to lead to LOF or an in-frame alteration and whether this type of effect is expected to be disease causing) (Abou Tayoun
108
+ et al.
109
+ 2018
110
+ 3
111
+ ).",Disease-specific
112
+ TNNT2 (HGNC:11949),PS3,Moderate,"In vivo models (e.g., variant knock-in animal models)
113
+
114
+
115
+ Mammalian variant-specific knock-in animal models that produce a phenotype consistent with the clinical phenotype in humans (e.g., structural and/or functional cardiac abnormalities, premature death, arrhythmia) may be considered as
116
+ MODERATE
117
+ evidence
118
+
119
+
120
+ NOTE:
121
+ The following assays/models do NOT meet criteria
122
+
123
+
124
+
125
+
126
+ Assays that are known to be associated with non-specific cardiac phenotypes (e.g., morpholino-induced pericardial edema in zebrafish)
127
+
128
+
129
+ In vivo evidence that is not variant specific, such as whole gene alterations (i.e., cDNA or whole gene transgenic mice and whole or partial gene knock-out mice)",Disease-specific
130
+ TNNT2 (HGNC:11949),PS3,Supporting,"In vitro
131
+
132
+ assays (e.g., biochemical assays of myofilament function, motility assays, human iPSC-CM)
133
+
134
+
135
+ While some
136
+ in vitro
137
+ assays may provide evidence that a variant in a cardiomyopathy gene has an effect on protein and/or myofilament function, at present, there are no validated “gold-standard” assays that are considered to reliably predict the clinical phenotype.
138
+
139
+
140
+ As such, in the cardiomyopathy genes listed in these guidelines, data from individual
141
+ in vitro
142
+ studies are unlikely to meet the criteria required to assign this rule at more than SUPPORTING level.",Disease-specific
143
+ TNNT2 (HGNC:11949),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
144
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
145
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
146
+ TNNT2 (HGNC:11949),PS4,Strong,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
147
+
148
+
149
+ Cohorts used in these analyses should meet the following criteria: 
150
+
151
+
152
+
153
+
154
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
155
+
156
+
157
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
158
+
159
+
160
+
161
+
162
+
163
+
164
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
165
+
166
+
167
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
168
+
169
+
170
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
171
+
172
+
173
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
174
+
175
+
176
+ The population diversity of the case and control cohorts are broadly similar.
177
+
178
+
179
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
180
+ et al.
181
+ 2017
182
+ 5
183
+ ,
184
+ DECIPHER
185
+ ).
186
+
187
+
188
+
189
+
190
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
191
+
192
+
193
+
194
+
195
+ STRONG
196
+ evidence requires the lower bound of the 95% confidence interval (CI) around the odds ratio (OR) estimate to be
197
+ ≥20
198
+
199
+
200
+
201
+
202
+ A PS4 calculator is available at
203
+ www.cardiodb.org
204
+ .
205
+
206
+
207
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
208
+
209
+
210
+ *RELEVANT PHENOTYPES:
211
+
212
+
213
+
214
+
215
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
216
+
217
+
218
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
219
+
220
+
221
+ Additional considerations for LVNC and end-stage HCM: 
222
+
223
+
224
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
225
+ et al.
226
+ 2017
227
+ 6
228
+ ; Oechslin
229
+ et al.
230
+ 2017
231
+ 7
232
+ ; Hershberger
233
+ et al.
234
+ 2017
235
+ 8
236
+ ; Ross
237
+ et al.
238
+ 2020
239
+ 9
240
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
241
+
242
+
243
+
244
+
245
+
246
+
247
+
248
+
249
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
250
+ TNNT2 (HGNC:11949),PS4,Moderate,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
251
+
252
+
253
+ Cohorts used in these analyses should meet the following criteria: 
254
+
255
+
256
+
257
+
258
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
259
+
260
+
261
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
262
+
263
+
264
+
265
+
266
+
267
+
268
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
269
+
270
+
271
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
272
+
273
+
274
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
275
+
276
+
277
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
278
+
279
+
280
+ The population diversity of the case and control cohorts are broadly similar.
281
+
282
+
283
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
284
+ et al.
285
+ 2017
286
+ 5
287
+ ,
288
+ DECIPHER
289
+ ).
290
+
291
+
292
+
293
+
294
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
295
+
296
+
297
+
298
+
299
+ MODERATE
300
+ evidence requires the lower bound of the 95% CI around the OR to be
301
+ ≥10
302
+
303
+
304
+
305
+
306
+ A PS4 calculator is available at
307
+ www.cardiodb.org
308
+ .
309
+
310
+
311
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
312
+
313
+
314
+ *RELEVANT PHENOTYPES:
315
+
316
+
317
+
318
+
319
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
320
+
321
+
322
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
323
+
324
+
325
+ Additional considerations for LVNC and end-stage HCM: 
326
+
327
+
328
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
329
+ et al.
330
+ 2017
331
+ 6
332
+ ; Oechslin
333
+ et al.
334
+ 2017
335
+ 7
336
+ ; Hershberger
337
+ et al.
338
+ 2017
339
+ 8
340
+ ; Ross
341
+ et al.
342
+ 2020
343
+ 9
344
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
345
+
346
+
347
+
348
+
349
+
350
+
351
+
352
+
353
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
354
+ TNNT2 (HGNC:11949),PS4,Supporting,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
355
+
356
+
357
+ Cohorts used in these analyses should meet the following criteria: 
358
+
359
+
360
+
361
+
362
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
363
+
364
+
365
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
366
+
367
+
368
+
369
+
370
+
371
+
372
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
373
+
374
+
375
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
376
+
377
+
378
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
379
+
380
+
381
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
382
+
383
+
384
+ The population diversity of the case and control cohorts are broadly similar.
385
+
386
+
387
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
388
+ et al.
389
+ 2017
390
+ 5
391
+ ,
392
+ DECIPHER
393
+ ).
394
+
395
+
396
+
397
+
398
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
399
+
400
+
401
+
402
+
403
+ SUPPORTING
404
+ evidence requires the lower bound of the 95% CI around the OR to be
405
+ ≥5
406
+
407
+
408
+
409
+
410
+ A PS4 calculator is available at
411
+ www.cardiodb.org
412
+ .
413
+
414
+
415
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
416
+
417
+
418
+ *RELEVANT PHENOTYPES:
419
+
420
+
421
+
422
+
423
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
424
+
425
+
426
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
427
+
428
+
429
+ Additional considerations for LVNC and end-stage HCM: 
430
+
431
+
432
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
433
+ et al.
434
+ 2017
435
+ 6
436
+ ; Oechslin
437
+ et al.
438
+ 2017
439
+ 7
440
+ ; Hershberger
441
+ et al.
442
+ 2017
443
+ 8
444
+ ; Ross
445
+ et al.
446
+ 2020
447
+ 9
448
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
449
+
450
+
451
+
452
+
453
+
454
+
455
+
456
+
457
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
458
+ TNNT2 (HGNC:11949),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
459
+ TNNT2 (HGNC:11949),PM1,Supporting,"Applicable to missense variants in
460
+ TNNT2
461
+ in the specific regions listed below (Walsh 
462
+ et al.
463
+ 2019
464
+ 10
465
+ ). 
466
+
467
+
468
+
469
+
470
+ Transcripts ENST00000367318 with codons 79-179
471
+
472
+
473
+ ENST00000656932.1 and NM_001276345.2 with codons 89-189
474
+
475
+
476
+
477
+
478
+ Data from HCM case cohorts was used to derive these cluster regions. Therefore, this rule should NOT be applied when additional evidence for the variant supports that the variant causes a phenotype other than HCM (e.g., variant seen in multiple DCM cases).
479
+
480
+
481
+ Enrichment was not observed for DCM in any genes.
482
+
483
+
484
+ Rule should NOT be combined with PM5 because presence of pathogenic variants in the same codon/region were used to determine clustering and would be double-counting evidence.","Disease-specific,Gene-specific"
485
+ TNNT2 (HGNC:11949),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
486
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
487
+ TNNT2 (HGNC:11949),PM2,Supporting,"The values used to calculate the PM2 thresholds were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/500 or lower), where the most frequent pathogenic variant accounts for no more than 2% of cases (e.g., has an allele frequency of ≤0.02 in cases based on the upper bound of 95% CI), and where the penetrance of a pathogenic variant is expected to be at least 50% (Kelly
488
+ et al.
489
+ 2018
490
+ 11
491
+ ).
492
+
493
+
494
+ A threshold of
495
+ ≤0.00004
496
+ in the subpopulation with the highest frequency when using the upper bound of the 95% CI activates this rule.
497
+
498
+
499
+
500
+
501
+ Alternatively, this is equivalent to the variant NOT being observed more than once (≤1 allele) in gnomAD v.2.1.1 in one of the non-founder populations (e.g., absence required from the Other and Ashkenazi Jewish subpopulations).
502
+
503
+
504
+ Applying a threshold of ≤0.00004 (upper bound of 95% CI of the allele frequency in gnomAD) is equivalent to the variant being seen in a single subpopulation and that subpopulation meets any of the following:
505
+
506
+
507
+ Allele Count (AC) in Allele Number (AN)
508
+
509
+
510
+ ≤1 in ≥120,000
511
+
512
+
513
+ ≤2 in ≥160,000
514
+
515
+
516
+ ≤3 in ≥195,000
517
+
518
+
519
+ ≤4 in ≥230,000
520
+
521
+
522
+
523
+
524
+
525
+
526
+
527
+
528
+ gnomAD is the preferred database for this calculation, but currently only displays the filtering allele frequency (FAF), which is equivalent to a lower bound estimate of the 95% CI, when the upper bound is what is needed.
529
+
530
+
531
+
532
+
533
+ Confidence interval tools, such as
534
+ Confit-de-MAF
535
+ , can be used to determine the upper bound of the 95% CI of the observed allele frequency.
536
+
537
+
538
+
539
+
540
+ Due to current technical limitations of next generation sequencing technologies, minor allele frequencies for complex variants (e.g., large indels) may not be accurately represented in population databases.
541
+
542
+
543
+ Caution should be used when a variant is only identified, or over-represented, in one of the smaller gnomAD populations, as the gnomAD allele frequencies may not accurately represent the true population frequency.
544
+
545
+
546
+ Population databases may contain affected or pre-symptomatic individuals for diseases with reduced penetrance/variable onset.",Disease-specific
547
+ TNNT2 (HGNC:11949),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
548
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
549
+ TNNT2 (HGNC:11949),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
550
+ TNNT2 (HGNC:11949),PM4,Moderate,"Strength of rule should be carefully considered and may require downgrading to SUPPORTING based on the predicted impact of the variant, including the size of the deletion/insertion, its location, and conservation of the region. 
551
+
552
+
553
+ For genes where PVS1 is not applicable (i.e., where there is no evidence that pLOF variants cause disease), consider using this rule at MODERATE or SUPPORTING strength for truncating variants that do NOT undergo nonsense mediated decay (NMD).",General recommendation
554
+ TNNT2 (HGNC:11949),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
555
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
556
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
557
+ TNNT2 (HGNC:11949),PM5,Moderate,"This criterion can be used at MODERATE if a different missense variant at the same codon has been classified as
558
+ pathogenic
559
+ using these modified guidelines without application of PM5.
560
+
561
+
562
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to SUPPORTING if the predicted impact is not expected to be equivalent or more severe.
563
+
564
+
565
+ PM5 should not be combined with PM1.  If both are applicable at MODERATE weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
566
+ TNNT2 (HGNC:11949),PM5,Supporting,"This criterion can be considered at SUPPORTING if a different missense variant at the same codon has been classified as
567
+ likely pathogenic
568
+ using these modified guidelines without application of PM5.
569
+
570
+
571
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as likely pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to NOT APPLICABLE if the predicted impact is not expected to be equivalent or more severe.
572
+
573
+
574
+ PM5 should not be combined with PM1.  The one with the higher strength should be applied, but if both are applicable at SUPPORTING weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
575
+ TNNT2 (HGNC:11949),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
576
+ TNNT2 (HGNC:11949),PM6,Moderate,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
577
+ 2
578
+ .
579
+
580
+
581
+ For most cardiomyopathies, it is recommended to default to “phenotype consistent with gene but not highly specific”. Clinical judgment is required for shifting to a higher or lower phenotypic consistency. 
582
+
583
+
584
+ See PS2 for additional considerations.",Disease-specific
585
+ TNNT2 (HGNC:11949),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
586
+ Note: May be used as stronger evidence with increasing segregation data.",
587
+ TNNT2 (HGNC:11949),PP1,Strong,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
588
+ ≥7
589
+
590
+ segregations
591
+ (LOD score of 2.1) for
592
+ STRONG
593
+ .
594
+
595
+
596
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
597
+ 12
598
+ can be considered: 
599
+
600
+
601
+
602
+
603
+ STRONG evidence requires ≥5 segregations (LOD score of 1.5)
604
+
605
+
606
+
607
+
608
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
609
+
610
+
611
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
612
+
613
+
614
+ Important considerations include:
615
+
616
+
617
+
618
+
619
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
620
+
621
+
622
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1.
623
+
624
+
625
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
626
+
627
+
628
+ Caution is needed when distantly related (≥3
629
+ rd
630
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
631
+ TNNT2 (HGNC:11949),PP1,Moderate,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
632
+ ≥5
633
+
634
+ segregations
635
+ (LOD score of 1.5) for
636
+ MODERATE
637
+ .
638
+
639
+
640
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
641
+ 12
642
+ can be considered: 
643
+
644
+
645
+
646
+
647
+ MODERATE evidence requires ≥4 segregations (LOD score of 1.2)
648
+
649
+
650
+
651
+
652
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
653
+
654
+
655
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
656
+
657
+
658
+ Important considerations include:
659
+
660
+
661
+
662
+
663
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
664
+
665
+
666
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
667
+
668
+
669
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
670
+
671
+
672
+ Caution is needed when distantly related (≥3
673
+ rd
674
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
675
+ TNNT2 (HGNC:11949),PP1,Supporting,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
676
+ ≥3
677
+
678
+ segregations
679
+ (LOD score of 0.9) for
680
+ SUPPORTING
681
+ . The thresholds as proposed by Jarvik and Browning (2016)
682
+ 12
683
+ are the same at ≥3 segregations (LOD score of 0.9) for supporting.
684
+
685
+
686
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
687
+
688
+
689
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
690
+
691
+
692
+ Important considerations include:
693
+
694
+
695
+
696
+
697
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
698
+
699
+
700
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
701
+
702
+
703
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
704
+
705
+
706
+ Caution is needed when distantly related (≥3
707
+ rd
708
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
709
+ TNNT2 (HGNC:11949),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
710
+ TNNT2 (HGNC:11949),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
711
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
712
+ TNNT2 (HGNC:11949),PP3,Supporting,"As many
713
+ in silico
714
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
715
+
716
+
717
+ Use of REVEL (Ioannidis
718
+ et al.
719
+ 2016
720
+ 13
721
+ ) is recommended at thresholds of
722
+ ≥0.70 for PP3
723
+ .
724
+
725
+
726
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
727
+
728
+
729
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
730
+
731
+
732
+ SpliceAI
733
+ 14
734
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
735
+ TNNT2 (HGNC:11949),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
736
+ TNNT2 (HGNC:11949),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
737
+ TNNT2 (HGNC:11949),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
738
+ TNNT2 (HGNC:11949),BA1,Stand Alone,"Allele frequency is
739
+ ≥0.001
740
+ based on the
741
+ filtering allele frequency (FAF)
742
+ in
743
+ gnomAD
744
+ in the subpopulation with the highest frequency (popmax).
745
+
746
+
747
+ The values used to calculate the BA1 threshold were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/300 or lower).
748
+
749
+
750
+ The threshold is applicable when assessing variants in the context of autosomal dominant cardiomyopathy. 
751
+
752
+
753
+ gnomAD is the preferred database for this calculation. If a subpopulation specific FAF other than the popmax is needed, this value can be calculated using the AlleleFrequencyApp on the
754
+ CardioDB website
755
+ .
756
+
757
+
758
+
759
+
760
+ Using the Inverse AF tab, enter in the population size and the number of alleles identified and it will calculate the FAF.  
761
+
762
+
763
+ Set confidence to 0.95 (95%).
764
+
765
+
766
+ If the FAF is ≥0.001, this rule can be applied.
767
+
768
+
769
+
770
+
771
+ The FAF by platform (e.g., exome vs. genome; v.2.1.1 vs. v.3.1.1) should be considered, the larger population is most likely to have the most accurate representation of “true” population allele frequency.
772
+
773
+
774
+ Caution is needed when considering any population cohorts that are smaller than the smallest subpopulations within gnomAD v.2.1.1 (e.g., ~5000 individuals or ~10,000 alleles). Despite this conservative nature of this threshold and approach, in smaller cohorts, the observed allele frequency may less accurately reflect the true allele frequency. Traditionally, once a variant is classified as Benign, it is rarely re-evaluated and so the highest confidence is needed to establish that classification on an allele frequency alone.",Disease-specific
775
+ TNNT2 (HGNC:11949),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
776
+ TNNT2 (HGNC:11949),BS1,Strong,"Allele frequency is
777
+ ≥0.0001 for
778
+
779
+ TNNT2
780
+ based on the
781
+ filtering allele frequency (FAF)
782
+ in
783
+ gnomAD
784
+ in the subpopulation with the highest frequency (popmax).
785
+
786
+
787
+ Criterion BS1 may only be used as standalone evidence to classify a variant as Likely Benign in the absence of conflicting data. See SVI guidance (Tavtigian
788
+ et al.
789
+ 2018
790
+ 15
791
+ ; Tavtigian
792
+ et al.
793
+ 2020
794
+ 16
795
+ ). 
796
+
797
+
798
+ See BA1 for additional specifications that also apply to BS1.","Disease-specific,Gene-specific"
799
+ TNNT2 (HGNC:11949),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",NA
800
+ TNNT2 (HGNC:11949),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
801
+ TNNT2 (HGNC:11949),BS3,Strong,See PS3 specifications.,Disease-specific
802
+ TNNT2 (HGNC:11949),BS3,Moderate,See PS3 specifications.,Disease-specific
803
+ TNNT2 (HGNC:11949),BS3,Supporting,See PS3 specifications.,Disease-specific
804
+ TNNT2 (HGNC:11949),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
805
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
806
+ TNNT2 (HGNC:11949),BS4,Strong,"Any non-segregations should be carefully evaluated to rule out a phenocopy or the presence of a second disease-causing variant before considering it as conflicting or benign evidence. 
807
+
808
+
809
+
810
+
811
+ The presence of “phenocopies” (e.g., athlete’s heart, hypertensive heart disease, ischemic cardiomyopathy, alcoholic cardiomyopathy, diabetic cardiomyopathy) can mimic non-segregation (i.e., lack of segregation) among affected individuals. 
812
+
813
+
814
+ Families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent ‘non-segregation’.
815
+
816
+
817
+
818
+
819
+ Because of these possibilities,
820
+ multiple (≥2) non-segregations
821
+ that are highly unlikely to be phenocopies or due to alternate variants (e.g., those without a possible alternate cause)
822
+ are required to apply this rule
823
+ .  A higher number of non-segregations is necessary for instances where alternative causes are possible (e.g., non-segregation in a sibling with childhood onset cardiomyopathy versus a grandparent with hypertension and HCM).
824
+
825
+
826
+ Careful consideration of the above points is required when using this data as conflicting evidence, especially when overall evidence supports likely pathogenic or pathogenic.",Disease-specific
827
+ TNNT2 (HGNC:11949),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
828
+ TNNT2 (HGNC:11949),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
829
+ TNNT2 (HGNC:11949),BP2,Supporting,"Other variants must be pathogenic as defined by these specifications.
830
+
831
+
832
+ Testing of parents or other informative relatives is often required to determine
833
+ cis
834
+ /
835
+ trans
836
+ status.
837
+
838
+
839
+ If a variant is seen in
840
+ trans
841
+ (or as double heterozygous) with another pathogenic variant in ≥2 cases and the phenotype is not more severe than when either of the two variants are seen in isolation, this rule may be applied (i.e., high confidence this variant is NOT contributing to disease).
842
+
843
+
844
+
845
+
846
+ <1% of cases of HCM have >1 pathogenic or likely pathogenic variant (0.6%; Alfares
847
+ et al.
848
+ 2015
849
+ 17
850
+ ).
851
+
852
+
853
+
854
+
855
+ This rule cannot be applied when the variant has only been observed in
856
+ cis
857
+ with a pathogenic variant as its significance in isolation is unknown in this scenario. 
858
+
859
+
860
+ Caution is needed if using this criterion as a primary piece of evidence for classifying a variant as likely benign/benign (i.e., only 2 SUPPORTING criteria are sufficient for a likely benign classification).",Disease-specific
861
+ TNNT2 (HGNC:11949),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
862
+ TNNT2 (HGNC:11949),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
863
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
864
+ TNNT2 (HGNC:11949),BP4,Supporting,"As many
865
+ in silico
866
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
867
+
868
+
869
+ Use of REVEL (Ioannidis et al. 2016
870
+ 13
871
+ ) is recommended at thresholds of
872
+ ≤0.40 for BP4
873
+ .
874
+
875
+
876
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
877
+
878
+
879
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
880
+
881
+
882
+ SpliceAI
883
+ 14
884
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
885
+ TNNT2 (HGNC:11949),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
886
+ TNNT2 (HGNC:11949),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
887
+ TNNT2 (HGNC:11949),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
888
+ TNNT2 (HGNC:11949),BP7,Supporting,"Also applicable to
889
+ intronic variants outside the splice consensus sequence (-4 and +7 outward)
890
+ for which splicing prediction algorithms predict no impact to the splice consensus sequence NOR the creation of a new splice site AND the nucleotide is not highly conserved.
891
+
892
+
893
+ Rule can be combined with BP4 to make a variant likely benign per Richards
894
+ et al.
895
+ 2015
896
+ 1
897
+ .",General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCardiomyopathyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforTPM1Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,898 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ TPM1 (HGNC:12010),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",NA
8
+ TPM1 (HGNC:12010),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
9
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
10
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
11
+ TPM1 (HGNC:12010),PS1,Strong,"No cardiomyopathy specifications. Apply as outlined by Richards
12
+ et al
13
+ . 2015
14
+ 1
15
+ .
16
+
17
+
18
+ Example of when rule should NOT be applied. NM_000256.3(
19
+ MYBPC3
20
+ ): c.2308G>A (p.Asp770Asn) has an established impact on splicing leading to nonsense mediated decay (NMD) and should not be used to provide evidence for other variants observed to result in the same amino acid change.",No change
21
+ TPM1 (HGNC:12010),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
22
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
23
+ TPM1 (HGNC:12010),PS2,Strong,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
24
+ 2
25
+ .
26
+
27
+
28
+ For most cardiomyopathies, it is recommended to default to
29
+ Phenotype consistency: “Phenotype consistent with gene but not highly specific”
30
+ . Clinical judgment is required for shifting to a higher or lower category. 
31
+
32
+
33
+ For use as a STRONG or VERY STRONG criterion, ideally parents have been thoroughly clinically evaluated without evidence of cardiomyopathy (ideally using a combination of ECG and echocardiogram or cardiac MRI for maximum sensitivity).
34
+
35
+
36
+ A family history consistent with
37
+ de novo
38
+ inheritance should not have any clinical signs or symptoms suggestive of cardiomyopathy in a 1
39
+ st
40
+ or 2
41
+ nd
42
+ degree relative, for example: 
43
+
44
+
45
+
46
+
47
+ Sudden death under 60 years of age
48
+
49
+
50
+ Heart transplant
51
+
52
+
53
+ Implantable cardiac defibrillator (ICD) under 60 years of age
54
+
55
+
56
+ Features of cardiomyopathy (e.g., systolic dysfunction, hypertrophy, left ventricular enlargement in an individual without risk factors).
57
+
58
+
59
+ Other related/overlapping cardiomyopathies
60
+
61
+
62
+
63
+
64
+ Examples of non-suspicious family history may include non-specific clinical features (e.g., palpitations, syncope, borderline/inconclusive echocardiogram findings, heart attack if age appropriate and suspected to result from coronary artery disease), but every attempt should be made to clarify features. 
65
+
66
+
67
+ Generally, this criterion is only applicable in the ABSENCE of any other possible disease-causing variants.  If other pathogenic or likely pathogenic variants are present, consider decreasing points assigned or overall weight.",Disease-specific
68
+ TPM1 (HGNC:12010),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
69
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
70
+ TPM1 (HGNC:12010),PS3,Strong,"In vitro splicing assays (e.g., RNA studies)
71
+
72
+
73
+ In vitro
74
+ splicing assays may be considered as 
75
+ STRONG
76
+ evidence, providing the following criteria are met.
77
+
78
+
79
+
80
+
81
+ Prior knowledge of predominant transcripts in cardiac tissue
82
+
83
+
84
+
85
+
86
+ Analysis undertaken using RNA extracted from cardiac tissue from the individual with the variant
87
+
88
+
89
+ Analysis undertaken using RNA extracted from whole blood providing the relevant transcripts (isoforms) are expressed in blood and are at sufficient levels to assess splice disruption.
90
+
91
+
92
+ Assay shows a clear, reproducible and convincing effect on splicing (i.e. a distinct splice product, present at a level comparable to the splice product from the wild-type allele), which is not observed in controls
93
+
94
+
95
+
96
+
97
+ Confirmation of abnormal splice product by Sanger sequencing
98
+
99
+
100
+
101
+
102
+ NOTE:
103
+ Mini-gene assay in non-patient derived cell lines are NOT considered to provide STRONG evidence.
104
+
105
+
106
+ NOTE:
107
+   Whether to activate this rule needs to be reconciled with the variant spectrum and disease mechanism for the gene at hand (i.e., consider whether the effect is likely to lead to LOF or an in-frame alteration and whether this type of effect is expected to be disease causing) (Abou Tayoun
108
+ et al.
109
+ 2018
110
+ 3
111
+ ).",Disease-specific
112
+ TPM1 (HGNC:12010),PS3,Moderate,"In vivo models (e.g., variant knock-in animal models)
113
+
114
+
115
+ Mammalian variant-specific knock-in animal models that produce a phenotype consistent with the clinical phenotype in humans (e.g., structural and/or functional cardiac abnormalities, premature death, arrhythmia) may be considered as
116
+ MODERATE
117
+ evidence
118
+
119
+
120
+ NOTE:
121
+ The following assays/models do NOT meet criteria
122
+
123
+
124
+
125
+
126
+ Assays that are known to be associated with non-specific cardiac phenotypes (e.g., morpholino-induced pericardial edema in zebrafish)
127
+
128
+
129
+ In vivo evidence that is not variant specific, such as whole gene alterations (i.e., cDNA or whole gene transgenic mice and whole or partial gene knock-out mice)",Disease-specific
130
+ TPM1 (HGNC:12010),PS3,Supporting,"In vitro
131
+
132
+ assays (e.g., biochemical assays of myofilament function, motility assays, human iPSC-CM)
133
+
134
+
135
+ While some
136
+ in vitro
137
+ assays may provide evidence that a variant in a cardiomyopathy gene has an effect on protein and/or myofilament function, at present, there are no validated “gold-standard” assays that are considered to reliably predict the clinical phenotype.
138
+
139
+
140
+ As such, in the cardiomyopathy genes listed in these guidelines, data from individual
141
+ in vitro
142
+ studies are unlikely to meet the criteria required to assign this rule at more than SUPPORTING level.",Disease-specific
143
+ TPM1 (HGNC:12010),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
144
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
145
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
146
+ TPM1 (HGNC:12010),PS4,Strong,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
147
+
148
+
149
+ Cohorts used in these analyses should meet the following criteria: 
150
+
151
+
152
+
153
+
154
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
155
+
156
+
157
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
158
+
159
+
160
+
161
+
162
+
163
+
164
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
165
+
166
+
167
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
168
+
169
+
170
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
171
+
172
+
173
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
174
+
175
+
176
+ The population diversity of the case and control cohorts are broadly similar.
177
+
178
+
179
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
180
+ et al.
181
+ 2017
182
+ 5
183
+ ,
184
+ DECIPHER
185
+ ).
186
+
187
+
188
+
189
+
190
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
191
+
192
+
193
+
194
+
195
+ STRONG
196
+ evidence requires the lower bound of the 95% confidence interval (CI) around the odds ratio (OR) estimate to be
197
+ ≥20
198
+
199
+
200
+
201
+
202
+ A PS4 calculator is available at
203
+ www.cardiodb.org
204
+ .
205
+
206
+
207
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
208
+
209
+
210
+ *RELEVANT PHENOTYPES:
211
+
212
+
213
+
214
+
215
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
216
+
217
+
218
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
219
+
220
+
221
+ Additional considerations for LVNC and end-stage HCM: 
222
+
223
+
224
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
225
+ et al.
226
+ 2017
227
+ 6
228
+ ; Oechslin
229
+ et al.
230
+ 2017
231
+ 7
232
+ ; Hershberger
233
+ et al.
234
+ 2017
235
+ 8
236
+ ; Ross
237
+ et al.
238
+ 2020
239
+ 9
240
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
241
+
242
+
243
+
244
+
245
+
246
+
247
+
248
+
249
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
250
+ TPM1 (HGNC:12010),PS4,Moderate,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
251
+
252
+
253
+ Cohorts used in these analyses should meet the following criteria: 
254
+
255
+
256
+
257
+
258
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
259
+
260
+
261
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
262
+
263
+
264
+
265
+
266
+
267
+
268
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
269
+
270
+
271
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
272
+
273
+
274
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
275
+
276
+
277
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
278
+
279
+
280
+ The population diversity of the case and control cohorts are broadly similar.
281
+
282
+
283
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
284
+ et al.
285
+ 2017
286
+ 5
287
+ ,
288
+ DECIPHER
289
+ ).
290
+
291
+
292
+
293
+
294
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
295
+
296
+
297
+
298
+
299
+ MODERATE
300
+ evidence requires the lower bound of the 95% CI around the OR to be
301
+ ≥10
302
+
303
+
304
+
305
+
306
+ A PS4 calculator is available at
307
+ www.cardiodb.org
308
+ .
309
+
310
+
311
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
312
+
313
+
314
+ *RELEVANT PHENOTYPES:
315
+
316
+
317
+
318
+
319
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
320
+
321
+
322
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
323
+
324
+
325
+ Additional considerations for LVNC and end-stage HCM: 
326
+
327
+
328
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
329
+ et al.
330
+ 2017
331
+ 6
332
+ ; Oechslin
333
+ et al.
334
+ 2017
335
+ 7
336
+ ; Hershberger
337
+ et al.
338
+ 2017
339
+ 8
340
+ ; Ross
341
+ et al.
342
+ 2020
343
+ 9
344
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
345
+
346
+
347
+
348
+
349
+
350
+
351
+
352
+
353
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
354
+ TPM1 (HGNC:12010),PS4,Supporting,"Currently few well-designed case-control studies have been performed for inherited cardiomyopathies.  Until such studies become available, comparative analyses can be undertaken using case data (e.g., internal and/or published cohorts) and control data from population-level cohorts (e.g., gnomAD). 
355
+
356
+
357
+ Cohorts used in these analyses should meet the following criteria: 
358
+
359
+
360
+
361
+
362
+ The cases have a clinical diagnosis of the specified disorder or related phenotype (e.g., all cases have HCM or another relevant phenotype*). 
363
+
364
+
365
+ When assessing cases, it's important to consider how likely another potential cause of the phenotype has been excluded.  This includes considering the presence of other variants in relevant genes (particularly those likely to be contributing to phenotype) and the extent of testing performed (i.e., single gene sequencing, panel testing, whole exome/genome sequencing).
366
+
367
+
368
+
369
+
370
+
371
+
372
+ The controls should not be derived from study populations that might be enriched for the specified disorder.
373
+
374
+
375
+ The denominator of the cohorts must be available (e.g., variant detected in 5 out of 3,500 cases and 1 out of 60,000 controls).
376
+
377
+
378
+ The cohorts do not include closely related individuals (i.e., family members are not included in the case counts).
379
+
380
+
381
+ The cohorts do not overlap with other cohorts being used in the analysis (i.e., cases are not being counted more than once).
382
+
383
+
384
+ The population diversity of the case and control cohorts are broadly similar.
385
+
386
+
387
+ Consider the size of the case cohort — larger cohorts are likely to provide more accurate estimates of variant frequency; therefore, it may be preferable to use data from the largest available case series for case-control analyses (e.g., Walsh
388
+ et al.
389
+ 2017
390
+ 5
391
+ ,
392
+ DECIPHER
393
+ ).
394
+
395
+
396
+
397
+
398
+ To account for limitations that arise when performing unmatched case-control analyses, the following stringent OR threshold is recommended:
399
+
400
+
401
+
402
+
403
+ SUPPORTING
404
+ evidence requires the lower bound of the 95% CI around the OR to be
405
+ ≥5
406
+
407
+
408
+
409
+
410
+ A PS4 calculator is available at
411
+ www.cardiodb.org
412
+ .
413
+
414
+
415
+ If multiple cohorts are available, the final ORs and associated CIs need to be harmonized across all cohorts to determine the final level (e.g., if 2 large cohorts have an OR of ~6 and a third small cohort has an OR of 11, application at a SUPPORTING level should be considered).  
416
+
417
+
418
+ *RELEVANT PHENOTYPES:
419
+
420
+
421
+
422
+
423
+ Cases of HCM and RCM may be combined as they are considered part of the same disease spectrum. 
424
+
425
+
426
+ For the eight genes covered by these guidelines, the combination of probands with other phenotypes should be reviewed by a clinical expert to determine if grouping is appropriate. 
427
+
428
+
429
+ Additional considerations for LVNC and end-stage HCM: 
430
+
431
+
432
+ Due to the current debate about whether isolated LVNC represents a true disease entity or variation of typical cardiac morphology (Anderson
433
+ et al.
434
+ 2017
435
+ 6
436
+ ; Oechslin
437
+ et al.
438
+ 2017
439
+ 7
440
+ ; Hershberger
441
+ et al.
442
+ 2017
443
+ 8
444
+ ; Ross
445
+ et al.
446
+ 2020
447
+ 9
448
+ ), individuals with isolated LVNC should NOT be added to proband or segregation counts (including individuals with isolated LVNC in a family with other cardiomyopathies).
449
+
450
+
451
+
452
+
453
+
454
+
455
+
456
+
457
+ HCM and DCM have distinct mechanisms of disease and therefore pathogenetic variants are not anticipated to cause both primary phenotypes. While occurrence in both phenotypes may initially be considered as evidence against pathogenicity, end-stage HCM can present similarly to DCM. Careful consideration is needed before including DCM or related phenotypes in case or segregation data for primarily HCM variants.",Disease-specific
458
+ TPM1 (HGNC:12010),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
459
+ TPM1 (HGNC:12010),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
460
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
461
+ TPM1 (HGNC:12010),PM2,Supporting,"The values used to calculate the PM2 thresholds were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/500 or lower), where the most frequent pathogenic variant accounts for no more than 2% of cases (e.g., has an allele frequency of ≤0.02 in cases based on the upper bound of 95% CI), and where the penetrance of a pathogenic variant is expected to be at least 50% (Kelly
462
+ et al.
463
+ 2018
464
+ 10
465
+ ).
466
+
467
+
468
+ A threshold of
469
+ ≤0.00004
470
+ in the subpopulation with the highest frequency when using the upper bound of the 95% CI activates this rule.
471
+
472
+
473
+
474
+
475
+ Alternatively, this is equivalent to the variant NOT being observed more than once (≤1 allele) in gnomAD v.2.1.1 in one of the non-founder populations (e.g., absence required from the Other and Ashkenazi Jewish subpopulations).
476
+
477
+
478
+ Applying a threshold of ≤0.00004 (upper bound of 95% CI of the allele frequency in gnomAD) is equivalent to the variant being seen in a single subpopulation and that subpopulation meets any of the following:
479
+
480
+
481
+ Allele Count (AC) in Allele Number (AN)
482
+
483
+
484
+ ≤1 in ≥120,000
485
+
486
+
487
+ ≤2 in ≥160,000
488
+
489
+
490
+ ≤3 in ≥195,000
491
+
492
+
493
+ ≤4 in ≥230,000
494
+
495
+
496
+
497
+
498
+
499
+
500
+
501
+
502
+ gnomAD is the preferred database for this calculation, but currently only displays the filtering allele frequency (FAF), which is equivalent to a lower bound estimate of the 95% CI, when the upper bound is what is needed.
503
+
504
+
505
+
506
+
507
+ Confidence interval tools, such as
508
+ Confit-de-MAF
509
+ , can be used to determine the upper bound of the 95% CI of the observed allele frequency.
510
+
511
+
512
+
513
+
514
+ Due to current technical limitations of next generation sequencing technologies, minor allele frequencies for complex variants (e.g., large indels) may not be accurately represented in population databases.
515
+
516
+
517
+ Caution should be used when a variant is only identified, or over-represented, in one of the smaller gnomAD populations, as the gnomAD allele frequencies may not accurately represent the true population frequency.
518
+
519
+
520
+ Population databases may contain affected or pre-symptomatic individuals for diseases with reduced penetrance/variable onset.",Disease-specific
521
+ TPM1 (HGNC:12010),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
522
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
523
+ TPM1 (HGNC:12010),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
524
+ TPM1 (HGNC:12010),PM4,Moderate,"Strength of rule should be carefully considered and may require downgrading to SUPPORTING based on the predicted impact of the variant, including the size of the deletion/insertion, its location, and conservation of the region. 
525
+
526
+
527
+ For genes where PVS1 is not applicable (i.e., where there is no evidence that pLOF variants cause disease), consider using this rule at MODERATE or SUPPORTING strength for truncating variants that do NOT undergo nonsense mediated decay (NMD).",General recommendation
528
+ TPM1 (HGNC:12010),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
529
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
530
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
531
+ TPM1 (HGNC:12010),PM5,Moderate,"This criterion can be used at MODERATE if a different missense variant at the same codon has been classified as
532
+ pathogenic
533
+ using these modified guidelines without application of PM5.
534
+
535
+
536
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to SUPPORTING if the predicted impact is not expected to be equivalent or more severe.
537
+
538
+
539
+ PM5 should not be combined with PM1.  If both are applicable at MODERATE weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
540
+ TPM1 (HGNC:12010),PM5,Supporting,"This criterion can be considered at SUPPORTING if a different missense variant at the same codon has been classified as
541
+ likely pathogenic
542
+ using these modified guidelines without application of PM5.
543
+
544
+
545
+ The impact of the amino acid change being evaluated needs to be compared to the impact of the amino acid change that is established as likely pathogenic (e.g., a change of Ala to His is less severe than Ala to Cys change). Consider reducing the strength of this rule to NOT APPLICABLE if the predicted impact is not expected to be equivalent or more severe.
546
+
547
+
548
+ PM5 should not be combined with PM1.  The one with the higher strength should be applied, but if both are applicable at SUPPORTING weight, use of PM5 is most appropriate since it is variant specific.",General recommendation
549
+ TPM1 (HGNC:12010),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
550
+ TPM1 (HGNC:12010),PM6,Moderate,"Refer to SVI guidance on number/combination of cases required based on phenotype specificity
551
+ 2
552
+ .
553
+
554
+
555
+ For most cardiomyopathies, it is recommended to default to “phenotype consistent with gene but not highly specific”. Clinical judgment is required for shifting to a higher or lower phenotypic consistency. 
556
+
557
+
558
+ See PS2 for additional considerations.",Disease-specific
559
+ TPM1 (HGNC:12010),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
560
+ Note: May be used as stronger evidence with increasing segregation data.",
561
+ TPM1 (HGNC:12010),PP1,Strong,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
562
+ ≥7
563
+
564
+ segregations
565
+ (LOD score of 2.1) for
566
+ STRONG
567
+ .
568
+
569
+
570
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
571
+ 11
572
+ can be considered: 
573
+
574
+
575
+
576
+
577
+ STRONG evidence requires ≥5 segregations (LOD score of 1.5)
578
+
579
+
580
+
581
+
582
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
583
+
584
+
585
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
586
+
587
+
588
+ Important considerations include:
589
+
590
+
591
+
592
+
593
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
594
+
595
+
596
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1.
597
+
598
+
599
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
600
+
601
+
602
+ Caution is needed when distantly related (≥3
603
+ rd
604
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
605
+ TPM1 (HGNC:12010),PP1,Moderate,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
606
+ ≥5
607
+
608
+ segregations
609
+ (LOD score of 1.5) for
610
+ MODERATE
611
+ .
612
+
613
+
614
+ Although rare for inherited cardiomyopathies, when the phenotype/presentation of a variant within and across families is highly specific (e.g., early-onset severe RCM in all affected individuals), the following thresholds as proposed by Jarvik and Browning (2016)
615
+ 11
616
+ can be considered: 
617
+
618
+
619
+
620
+
621
+ MODERATE evidence requires ≥4 segregations (LOD score of 1.2)
622
+
623
+
624
+
625
+
626
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
627
+
628
+
629
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
630
+
631
+
632
+ Important considerations include:
633
+
634
+
635
+
636
+
637
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
638
+
639
+
640
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
641
+
642
+
643
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
644
+
645
+
646
+ Caution is needed when distantly related (≥3
647
+ rd
648
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
649
+ TPM1 (HGNC:12010),PP1,Supporting,"Due to the genotypic and phenotypic heterogeneity of inherited cardiomyopathies, segregation thresholds have been conservatively set at
650
+ ≥3
651
+
652
+ segregations
653
+ (LOD score of 0.9) for
654
+ SUPPORTING
655
+ . The thresholds as proposed by Jarvik and Browning (2016)
656
+ 11
657
+ are the same at ≥3 segregations (LOD score of 0.9) for supporting.
658
+
659
+
660
+ Only genotype positive/phenotype positive individuals are counted as segregations, which can include affected obligate carriers. Genotype positive/phenotype negative individuals are generally less informative for cardiomyopathy genes due to variable age at onset and reduced penetrance.
661
+
662
+
663
+ Phenotypes should be clinically confirmed, whenever possible, and should not include individuals with a suspected diagnosis.  
664
+
665
+
666
+ Important considerations include:
667
+
668
+
669
+
670
+
671
+ Segregation of a variant within a single family or haplotype has the potential to represent linkage disequilibrium with another undetected variant.  If linkage disequilibrium is a concern, consider downgrading strength of segregation. 
672
+
673
+
674
+ Use of segregation criteria should be carefully evaluated if variant frequency meets criteria for BS1 (see below).
675
+
676
+
677
+ Caution is needed when counting segregations in presence of other possible disease-causing variants, as both variants may be contributing to the phenotype. 
678
+
679
+
680
+ Caution is needed when distantly related (≥3
681
+ rd
682
+ degree) affected individuals are connected by unknown or unaffected relatives (raises possibility of multiple causes of disease).",Disease-specific
683
+ TPM1 (HGNC:12010),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,
684
+ TPM1 (HGNC:12010),PP2,Supporting,"Application of this rule takes into consideration empirical data quantifying levels of rare missense variant enrichment in HCM referral cohorts compared to population-based cohorts (Walsh
685
+ et al.
686
+ 2019
687
+ 12
688
+ ) rather than the missense constraint score in gnomAD. 
689
+
690
+
691
+ On the basis of data from Walsh
692
+ et al.
693
+ 2019
694
+ 12
695
+ ,
696
+ PP2 is currently
697
+
698
+ only applicable to
699
+
700
+ TPM1
701
+
702
+ for HCM
703
+ (transcripts ENST00000403994 and NM_001018005.2)
704
+ .
705
+
706
+
707
+ Data from HCM case cohorts was used to derive these cluster regions. Therefore, this rule should NOT be applied when additional evidence for the variant supports that the variant causes a phenotype other than HCM (e.g., variant seen in multiple DCM cases).
708
+
709
+
710
+ Enrichment was not observed for DCM in any genes.","Disease-specific,Gene-specific"
711
+ TPM1 (HGNC:12010),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
712
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
713
+ TPM1 (HGNC:12010),PP3,Supporting,"As many
714
+ in silico
715
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
716
+
717
+
718
+ Use of REVEL (Ioannidis
719
+ et al.
720
+ 2016
721
+ 13
722
+ ) is recommended at thresholds of
723
+ ≥0.70 for PP3
724
+ .
725
+
726
+
727
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
728
+
729
+
730
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
731
+
732
+
733
+ SpliceAI
734
+ 14
735
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
736
+ TPM1 (HGNC:12010),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
737
+ TPM1 (HGNC:12010),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
738
+ TPM1 (HGNC:12010),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
739
+ TPM1 (HGNC:12010),BA1,Stand Alone,"Allele frequency is
740
+ ≥0.001
741
+ based on the
742
+ filtering allele frequency (FAF)
743
+ in
744
+ gnomAD
745
+ in the subpopulation with the highest frequency (popmax).
746
+
747
+
748
+ The values used to calculate the BA1 threshold were derived from studies in Northern European populations that have been relatively well-characterized with regards to disease prevalence and variant spectrum. These thresholds can be applied to any population where disease prevalence is considered comparable (1/300 or lower).
749
+
750
+
751
+ The threshold is applicable when assessing variants in the context of autosomal dominant cardiomyopathy. 
752
+
753
+
754
+ gnomAD is the preferred database for this calculation. If a subpopulation specific FAF other than the popmax is needed, this value can be calculated using the AlleleFrequencyApp on the
755
+ CardioDB website
756
+ .
757
+
758
+
759
+
760
+
761
+ Using the Inverse AF tab, enter in the population size and the number of alleles identified and it will calculate the FAF.  
762
+
763
+
764
+ Set confidence to 0.95 (95%).
765
+
766
+
767
+ If the FAF is ≥0.001, this rule can be applied.
768
+
769
+
770
+
771
+
772
+ The FAF by platform (e.g., exome vs. genome; v.2.1.1 vs. v.3.1.1) should be considered, the larger population is most likely to have the most accurate representation of “true” population allele frequency.
773
+
774
+
775
+ Caution is needed when considering any population cohorts that are smaller than the smallest subpopulations within gnomAD v.2.1.1 (e.g., ~5000 individuals or ~10,000 alleles). Despite this conservative nature of this threshold and approach, in smaller cohorts, the observed allele frequency may less accurately reflect the true allele frequency. Traditionally, once a variant is classified as Benign, it is rarely re-evaluated and so the highest confidence is needed to establish that classification on an allele frequency alone.",Disease-specific
776
+ TPM1 (HGNC:12010),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
777
+ TPM1 (HGNC:12010),BS1,Strong,"Allele frequency is
778
+ ≥0.0001 for
779
+
780
+ TPM1
781
+ based on the
782
+ filtering allele frequency (FAF)
783
+ in
784
+ gnomAD
785
+ in the subpopulation with the highest frequency (popmax).
786
+
787
+
788
+ Criterion BS1 may only be used as standalone evidence to classify a variant as Likely Benign in the absence of conflicting data. See SVI guidance (Tavtigian
789
+ et al.
790
+ 2018
791
+ 15
792
+ ; Tavtigian
793
+ et al.
794
+ 2020
795
+ 16
796
+ ). 
797
+
798
+
799
+ See BA1 for additional specifications that also apply to BS1.","Disease-specific,Gene-specific"
800
+ TPM1 (HGNC:12010),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",NA
801
+ TPM1 (HGNC:12010),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
802
+ TPM1 (HGNC:12010),BS3,Strong,See PS3 specifications.,Disease-specific
803
+ TPM1 (HGNC:12010),BS3,Moderate,See PS3 specifications.,Disease-specific
804
+ TPM1 (HGNC:12010),BS3,Supporting,See PS3 specifications.,Disease-specific
805
+ TPM1 (HGNC:12010),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
806
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
807
+ TPM1 (HGNC:12010),BS4,Strong,"Any non-segregations should be carefully evaluated to rule out a phenocopy or the presence of a second disease-causing variant before considering it as conflicting or benign evidence. 
808
+
809
+
810
+
811
+
812
+ The presence of “phenocopies” (e.g., athlete’s heart, hypertensive heart disease, ischemic cardiomyopathy, alcoholic cardiomyopathy, diabetic cardiomyopathy) can mimic non-segregation (i.e., lack of segregation) among affected individuals. 
813
+
814
+
815
+ Families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent ‘non-segregation’.
816
+
817
+
818
+
819
+
820
+ Because of these possibilities,
821
+ multiple (≥2) non-segregations
822
+ that are highly unlikely to be phenocopies or due to alternate variants (e.g., those without a possible alternate cause)
823
+ are required to apply this rule
824
+ .  A higher number of non-segregations is necessary for instances where alternative causes are possible (e.g., non-segregation in a sibling with childhood onset cardiomyopathy versus a grandparent with hypertension and HCM).
825
+
826
+
827
+ Careful consideration of the above points is required when using this data as conflicting evidence, especially when overall evidence supports likely pathogenic or pathogenic.",Disease-specific
828
+ TPM1 (HGNC:12010),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
829
+ TPM1 (HGNC:12010),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
830
+ TPM1 (HGNC:12010),BP2,Supporting,"Other variants must be pathogenic as defined by these specifications.
831
+
832
+
833
+ Testing of parents or other informative relatives is often required to determine
834
+ cis
835
+ /
836
+ trans
837
+ status.
838
+
839
+
840
+ If a variant is seen in
841
+ trans
842
+ (or as double heterozygous) with another pathogenic variant in ≥2 cases and the phenotype is not more severe than when either of the two variants are seen in isolation, this rule may be applied (i.e., high confidence this variant is NOT contributing to disease).
843
+
844
+
845
+
846
+
847
+ <1% of cases of HCM have >1 pathogenic or likely pathogenic variant (0.6%; Alfares
848
+ et al.
849
+ 2015
850
+ 17
851
+ ).
852
+
853
+
854
+
855
+
856
+ This rule cannot be applied when the variant has only been observed in
857
+ cis
858
+ with a pathogenic variant as its significance in isolation is unknown in this scenario. 
859
+
860
+
861
+ Caution is needed if using this criterion as a primary piece of evidence for classifying a variant as likely benign/benign (i.e., only 2 SUPPORTING criteria are sufficient for a likely benign classification).",Disease-specific
862
+ TPM1 (HGNC:12010),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
863
+ TPM1 (HGNC:12010),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
864
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
865
+ TPM1 (HGNC:12010),BP4,Supporting,"As many
866
+ in silico
867
+ algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. Meta-predictors, such as REVEL, are preferred over multiple individual predictors.
868
+
869
+
870
+ Use of REVEL (Ioannidis et al. 2016
871
+ 13
872
+ ) is recommended at thresholds of
873
+ ≤0.40 for BP4
874
+ .
875
+
876
+
877
+ Clinical judgment is needed if any individual algorithms or conservation data are contradictory to REVEL data.
878
+
879
+
880
+ Positive predictive value for benign/no impact predictions is generally higher than for pathogenic/impact predictions.
881
+
882
+
883
+ SpliceAI
884
+ 14
885
+ is recommended for evaluation of predicted splice impacts.",Disease-specific
886
+ TPM1 (HGNC:12010),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
887
+ TPM1 (HGNC:12010),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
888
+ TPM1 (HGNC:12010),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
889
+ TPM1 (HGNC:12010),BP7,Supporting,"Also applicable to
890
+ intronic variants outside the splice consensus sequence (-4 and +7 outward)
891
+ for which splicing prediction algorithms predict no impact to the splice consensus sequence NOR the creation of a new splice site AND the nucleotide is not highly conserved.
892
+
893
+
894
+ Rule can be combined with BP4 to make a variant likely benign per Richards
895
+ et al.
896
+ 2015
897
+ 1
898
+ .",General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGAMTVersion1.1.0_version=1.1.0.csv ADDED
@@ -0,0 +1,324 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ GAMT (HGNC:4136),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ GAMT (HGNC:4136),PVS1,Very Strong,"Nonsense-mediated decay predicted
9
+
10
+
11
+ CCDS VCEP notes:
12
+ Loss of function (LOF) of GAMT is a known mechanism of disease for guanidinoacetate methyltransferase deficiency (GAMT-D). There are examples of various LOF variants, including nonsense and frameshift, in GAMT in individuals with GAMT-D (
13
+ https://databases.lovd.nl/shared/variants/GAMT/unique
14
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
15
+
16
+
17
+ Nonsense and frameshift variants
18
+ * All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 6) or the last 50 bases of the penultimate exon of the gene (exon 5, c.520). In that case, PVS1_Strong or PVS1_Moderate will be applied, depending on whether >10% or <10% of the protein is lost.
19
+
20
+
21
+ Splice site variants (+1, +2, -1, -2)
22
+ * All canonical splice site pairs in GAMT are GT-AG.
23
+ * For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
24
+ * For the predicted in frame/out of frame consequences of exon skipping and assigned strength of PVS1, see
25
+ Appendix 1
26
+ .
27
+ * If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
28
+ * To apply PVS1, splice site variants must have no detectable nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. Otherwise, the PVS1 strength should be reduced accordingly.
29
+ * Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
30
+
31
+
32
+ Deletions (single or multi exon)
33
+ * If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
34
+ * If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
35
+ * If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
36
+ *
37
+ Appendix 1
38
+ can be used to predict the consequences of single exon deletions.
39
+
40
+
41
+ Duplications
42
+ * Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",None
43
+ GAMT (HGNC:4136),PVS1,Strong,"In frame loss of >10% of the protein.
44
+
45
+
46
+ CCDS VCEP notes:
47
+ Loss of function (LOF) of GAMT is a known mechanism of disease for guanidinoacetate methyltransferase deficiency (GAMT-D). There are examples of various LOF variants, including nonsense and frameshift, in GAMT in individuals with GAMT-D (
48
+ https://databases.lovd.nl/shared/variants/GAMT/unique
49
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
50
+
51
+
52
+ Nonsense and frameshift variants
53
+ * All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 6) or the last 50 bases of the penultimate exon of the gene (exon 5, c.520). In that case, PVS1_Strong will be applied if >10% of the protein is lost.
54
+
55
+
56
+ Splice site variants (+1, +2, -1, -2)
57
+ * All canonical splice site pairs in GAMT are GT-AG.
58
+ * For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
59
+ * Use SpliceAI and varSEAK to look for nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. 
60
+ * For considerations for strength at which PVS1 may be applied see
61
+ Appendix 1
62
+ .
63
+ * If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
64
+ * Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
65
+
66
+
67
+ Deletions (single or multi exon)
68
+ * If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
69
+ * If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
70
+ * If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
71
+ *
72
+ Appendix 1
73
+ can be used to predict the consequences of single exon deletions.
74
+
75
+
76
+ Duplications
77
+ * Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",Strength
78
+ GAMT (HGNC:4136),PVS1,Moderate,"Single exon or larger deletion resulting in loss of <10% of the protein.
79
+
80
+
81
+ Initiator codon variant.
82
+  
83
+
84
+
85
+ CCDS VCEP notes:
86
+ Loss of function (LOF) of GAMT is a known mechanism of disease for guanidinoacetate methyltransferase deficiency (GAMT-D). There are examples of various LOF variants, including nonsense and frameshift, in GAMT in individuals with GAMT-D (
87
+ https://databases.lovd.nl/shared/variants/GAMT/unique
88
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
89
+
90
+
91
+ Nonsense and frameshift variants
92
+ * All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 6) or the last 50 bases of the penultimate exon of the gene (exon 5, c.520). In that case, PVS1_Moderate will be applied if <10% of the protein is lost.
93
+
94
+
95
+ Splice site variants (+1, +2, -1, -2)
96
+ * All canonical splice site pairs in GAMT are GT-AG.
97
+ * Use SpliceAI and varSEAK to look for nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. 
98
+ * For any canonical splice site variant (+1, +2, -1, -2), if RT-PCR data predicts in-frame loss of <10% of the protein, apply PVS1_Moderate.
99
+ * If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
100
+ * Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
101
+
102
+
103
+ Initiator codon variants
104
+ * To our knowledge, initiator codon variants have not been reported in GAMT (01/2019) but may occur.
105
+ * All initiator codon variants will meet PVS1_Moderate. The next in-frame methionine is at amino acid position 42 (based on NP_000147.1).
106
+
107
+
108
+ Deletions (single or multi exon)
109
+ * If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
110
+ * If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
111
+ * If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
112
+ *
113
+ Appendix 1
114
+ can be used to predict the consequences of single exon deletions.
115
+
116
+
117
+ Duplications
118
+ * Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",Strength
119
+ GAMT (HGNC:4136),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
120
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
121
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
122
+ GAMT (HGNC:4136),PS1,Strong,"This criterion is applicable as described.
123
+
124
+
125
+ If the variant is in the last 3 nucleotides of an exon, further analysis using splicing site prediction algorithms (see PP3) and data from the literature (if available) is required to investigate the impact on splicing.",None
126
+ GAMT (HGNC:4136),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
127
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
128
+ GAMT (HGNC:4136),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
129
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
130
+ GAMT (HGNC:4136),PS3,Strong,"RT-PCR evidence of mis-splicing for non-canonical +1, +2, -2, -1 intronic variants with no evidence of normal splice products.
131
+  
132
+
133
+
134
+ GAMT specifications:
135
+ Any variant meeting the description for splicing assays below can meet PS3 or PS3_Supporting, and variants meeting the description for in vitro activity assays can meet PS3_Supporting. If a variant meets the description for both e.g., a splice site variant with evidence of abnormal splicing and deficient GAMT activity in vitro, PS3 must only be counted once.
136
+
137
+
138
+ Splicing assays
139
+ For non-canonical splicing variants, use PS3 if there is RT-PCR and/or RNA sequencing evidence demonstrating only abnormal splice products, with no evidence of normal splicing.
140
+ * Evidence of abnormal splicing includes transcripts of alternative length or with specific intron or exon inclusion/exclusion. These studies can be performed on mRNA extracted from patient-derived cells, or by inserting the mutant genomic DNA into plasmid vectors and introducing these into human or other mammalian host cells.
141
+ * Note that in patients who are compound heterozygotes for a splicing variant and another variants type that does not disrupt splicing, such as a missense variant, evidence of normal splicing is expected. However, the presence of normal splice products could complicate the assessment of the impact of the splice variant. Therefore, if there is any evidence of normal splice products, either when using RNA from patient cells or in an in vitro expression system, use PS3_Supporting.
142
+ * PP3 may also be used for non-canonical splice variants meeting PS3 or PS3_Supporting.",Disease-specific
143
+ GAMT (HGNC:4136),PS3,Supporting,"<15% control activity when variant is expressed in GAMT-deficient fibroblasts or HeLa cells.
144
+ * RT-PCR evidence of mis-splicing for non-canonical intronic variants with evidence of normal splice products.
145
+  
146
+
147
+
148
+ GAMT specifications:
149
+ Any variant meeting the description for splicing assays below can meet PS3 or PS3_Supporting, and variants meeting the description for in vitro activity assays can meet PS3_Supporting. If a variant meets the description for both e.g., a splice site variant with evidence of abnormal splicing and deficient GAMT activity in vitro, PS3 must only be counted once.
150
+
151
+
152
+ In vitro expression
153
+ PS3_Supporting can be assigned if a variant is expressed in GAMT-deficient fibroblasts or HeLa cells and has <15% of the control value, for values published in Mercimek-Mahmutoglu et al, 2014, PMID 24415674; Mercimek-Mahmutoglu et al, 2016, PMID 26319512; or DesRoches et al, 2016, PMID 26003046. See
154
+ Appendix 2
155
+ for further details on GAMT functional assays.  
156
+
157
+
158
+ Splicing assays
159
+ For non-canonical splicing variants, use PS3 if there is RT-PCR and/or RNA sequencing evidence demonstrating only abnormal splice products, with no evidence of normal splicing.
160
+ * Evidence of abnormal splicing includes transcripts of alternative length or with specific intron or exon inclusion/exclusion. These studies can be performed on mRNA extracted from patient-derived cells, or by inserting the mutant genomic DNA into plasmid vectors and introducing these into human or other mammalian host cells.
161
+ * Note that in patients who are compound heterozygotes for a splicing variant and another variants type that does not disrupt splicing, such as a missense variant, evidence of normal splicing is expected. However, the presence of normal splice products could complicate the assessment of the impact of the splice variant. Therefore, if there is any evidence of normal splice products, either when using RNA from patient cells or in an in vitro expression system, use PS3_Supporting.
162
+ * PP3 may also be used for non-canonical splice variants meeting PS3 or PS3_Supporting.",Disease-specific
163
+ GAMT (HGNC:4136),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
164
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
165
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",NA
166
+ GAMT (HGNC:4136),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
167
+ GAMT (HGNC:4136),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
168
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
169
+ GAMT (HGNC:4136),PM2,Supporting,"Allele frequency <0.0004 (<0.04%) in all populations in gnomAD.
170
+
171
+
172
+ CCDS VCEP notes: It is acceptable for a GAMT variant to be present in controls, if heterozygous, because GAMT-D is a recessive disorder. Homozygotes should not be seen in a population database, such as gnomAD, because the penetrance of this condition in individuals with biallelic pathogenic variants is expected to be 100%.
173
+
174
+
175
+ GAMT specifications:
176
+
177
+
178
+
179
+
180
+ All subpopulations in gnomAD must have a maximum allele frequency less than 0.0004 (the highest population minor allele frequency of the most common pathogenic GAMT variant, c.327G>A, in gnomAD). Any variant with a frequency below this cutoff will meet PM2_Supporting (See Appendix 3). Note – PM2 will NOT be used at moderate strength; PM2 will only be applied as a Supporting criterion.
181
+
182
+
183
+ If homozygotes are observed, the variant will meet BS2 (assuming 100% penetrance for an individual with 2 pathogenic variants in trans).","Disease-specific,Strength"
184
+ GAMT (HGNC:4136),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
185
+ Note: This requires testing of parents (or offspring) to determine phase.",
186
+ GAMT (HGNC:4136),PM3,Very Strong,"Follow SVI guidance for PM3 (
187
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
188
+ ).
189
+
190
+
191
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
192
+ GAMT (HGNC:4136),PM3,Strong,"Follow SVI guidance for PM3 (
193
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
194
+ ).
195
+
196
+
197
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
198
+ GAMT (HGNC:4136),PM3,Moderate,"Follow SVI guidance for PM3 (
199
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
200
+ ).
201
+
202
+
203
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",None
204
+ GAMT (HGNC:4136),PM3,Supporting,"Follow SVI guidance for PM3 (
205
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
206
+ ).
207
+
208
+
209
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
210
+ GAMT (HGNC:4136),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
211
+ GAMT (HGNC:4136),PM4,Moderate,"Stop loss variants in GAMT have not been reported, as far as we are aware.
212
+
213
+
214
+ GAMT specifications: Use this rule “as is” for in frame deletions and insertions of 2 or more amino acids, but downgrade to PM4_Supporting for single amino acid deletions and insertions.",None
215
+ GAMT (HGNC:4136),PM4,Supporting,Downgrade to PM4_Supporting for single amino acid deletions and insertions.,Strength
216
+ GAMT (HGNC:4136),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
217
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
218
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
219
+ GAMT (HGNC:4136),PM5,Moderate,"If the pathogenicity of another missense change at the same amino acid residue is unknown, determine its pathogenicity using these specifications in order to determine if this criterion can be used. If the variant is pathogenic, use PM5. If the variant is likely pathogenic, use PM5_Supporting.",None
220
+ GAMT (HGNC:4136),PM5,Supporting,"If the pathogenicity of another missense change at the same amino acid residue is unknown, determine its pathogenicity using these specifications in order to determine if this criterion can be used. If the variant is pathogenic, use PM5. If the variant is likely pathogenic, use PM5_Supporting.",Strength
221
+ GAMT (HGNC:4136),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
222
+ GAMT (HGNC:4136),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
223
+ Note: May be used as stronger evidence with increasing segregation data.",NA
224
+ GAMT (HGNC:4136),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
225
+ GAMT (HGNC:4136),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
226
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
227
+ GAMT (HGNC:4136),PP3,Moderate,"Non-canonical splice site variant predicted to be more deleterious, by SpliceAI and varSEAK, than a previously observed pathogenic variant at the same nucleotide.",Strength
228
+ GAMT (HGNC:4136),PP3,Supporting,"For missense changes, those with a REVEL score more than 0.75 will meet PP3.
229
+
230
+
231
+ For in frame insertions and deletions, use PROVEAN and Mutation Taster. Results must be consistent to count.
232
+
233
+
234
+ For non-canonical splice site variants (e.g., +3, -3), use SpliceAI (
235
+ https://spliceailookup.broadinstitute.org/
236
+ ) and varSEAK (
237
+ https://varseak.bio/
238
+ ). Results must be consistent to apply this criterion.
239
+
240
+
241
+ For SpliceAI, any donor loss or acceptor loss with a score >0.5. For varSEAK, any variant with splicing class 4 or 5. Evidence for creation of a cryptic splice site should also be assessed.
242
+
243
+
244
+ Do not apply this rule for canonical splice site changes meeting PVS1.",None
245
+ GAMT (HGNC:4136),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
246
+ GAMT (HGNC:4136),PP4,Strong,"4 points based on any combination of the following. Two or more data types are required to apply PP4_Strong:
247
+ • Elevated urine guanidinoacetate with or without low or low normal creatine (1 point)
248
+ • Elevated plasma guanidinoacetate with or without low or low normal creatine (2 points)
249
+ • Significantly decreased creatine peak in brain magnetic resonance spectroscopy with or without visible guanidinoacetate peak (3 points)
250
+ • GAMT enzyme activity <5% of normal (3 points)
251
+
252
+
253
+ * Variant must meet PM2_Supporting for PP4 to apply at any strength.
254
+ * For PP4 to be applied at strong, full GAMT gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
255
+ GAMT (HGNC:4136),PP4,Moderate,"3 points based on any combination of the following. Two or more data types are recommended to apply PP4_Moderate:
256
+ • Elevated urine guanidinoacetate with or without low or low normal creatine (1 point)
257
+ • Elevated plasma guanidinoacetate with or without low or low normal creatine (2 points)
258
+ • Significantly decreased creatine peak in brain magnetic resonance spectroscopy with or without visible guanidinoacetate peak (3 points)
259
+ • GAMT enzyme activity <5% of normal (3 points)
260
+
261
+
262
+ * Variant must meet PM2_Supporting for PP4 to apply at any strength.",Strength
263
+ GAMT (HGNC:4136),PP4,Supporting,"1-2 points based on: 
264
+ • Elevated urine guanidinoacetate with or without low or low normal creatine (1 point)
265
+ • Elevated plasma guanidinoacetate with or without low or low normal creatine (2 points)
266
+
267
+
268
+ * Variant must meet PM2_Supporting for PP4 to apply at any strength",Disease-specific
269
+ GAMT (HGNC:4136),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
270
+ GAMT (HGNC:4136),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
271
+ GAMT (HGNC:4136),BA1,Stand Alone,"Allele frequency >0.003 (0.3%) in gnomAD in any continental population in gnomAD with >2000 alleles.
272
+
273
+
274
+ GAMT specifications:
275
+
276
+
277
+
278
+
279
+ Any variant with a frequency >0.003 (based on the estimated prevalence 1 in 114,000, PMID 24071436) in gnomAD (max allelic contribution = 100%; max genetic contribution = 100%).
280
+
281
+
282
+ Use the highest population minor allele frequency (MAF) in any given continental population with >2,000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383).",Disease-specific
283
+ GAMT (HGNC:4136),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
284
+ GAMT (HGNC:4136),BS1,Strong,"Allele frequency >0.001 (0.1%) in gnomAD in any continental population in gnomAD with >2000 alleles. 
285
+
286
+
287
+ GAMT specifications:
288
+
289
+
290
+
291
+
292
+ Any variant with a frequency >0.001 (based on the estimated prevalence 1 in 114,000 (PMID: 24071436) in gnomAD (max allelic contribution = 40%; max genetic contribution = 100%).
293
+
294
+
295
+ Use the highest population minor allele frequency (MAF) in any given continental population with >2,000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383).",Disease-specific
296
+ GAMT (HGNC:4136),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
297
+ GAMT (HGNC:4136),BS2,Strong,Observed in the homozygous state in a healthy adult.,None
298
+ GAMT (HGNC:4136),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
299
+ GAMT (HGNC:4136),BS3,Supporting,">30% normal GAA activity when the variant is expressed in a heterologous cell type.
300
+ GAMT specifications:
301
+ In vitro assays in which a variant is expressed in GAMT-deficient cultured cells (e.g. GAMT-deficient fibroblasts) or in-fusion High-Fidelity cloning of GAMT transcript and site directed mutagenesis to generate missense variant overexpressed in HeLa cells and measurement of GAMT activity in cells for wild-type and missense variant. Any variant with enzyme activity at or above 30% of normal in the following publications meets BS3_Supporting (Mercimek-Mahmutoglu et al, 2014; PMID 24415674; Mercimek-Mahmutoglu et al, 2016, PMID 26319512; DesRoches et al, 2016, PMID 26003046).","Disease-specific,Strength"
302
+ GAMT (HGNC:4136),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
303
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
304
+ GAMT (HGNC:4136),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
305
+ GAMT (HGNC:4136),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
306
+ GAMT (HGNC:4136),BP2,Supporting,Observed in cis with a pathogenic variant (to take AR inheritance into account).,None
307
+ GAMT (HGNC:4136),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
308
+ GAMT (HGNC:4136),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
309
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
310
+ GAMT (HGNC:4136),BP4,Supporting,"For missense changes, REVEL score <0.5.
311
+
312
+
313
+ For in frame insertions and deletions, use PROVEAN and Mutation Taster. Results must be consistent to count.
314
+
315
+
316
+ For non-canonical splice site variants, use SpliceAI (
317
+ https://spliceailookup.broadinstitute.org/
318
+ ) and varSEAK (
319
+ https://varseak.bio/
320
+ ) to assess the impact of variants that are not +/-1 or 2 canonical splice site variants. For SpliceAI, this criterion can be applied for scores <0.2, and for varSEAK class 1 and 2. If there is any evidence for possible creation of a cryptic splice site, this criterion should not be applied.",None
321
+ GAMT (HGNC:4136),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
322
+ GAMT (HGNC:4136),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
323
+ GAMT (HGNC:4136),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
324
+ GAMT (HGNC:4136),BP7,Supporting,Apply this criterion as described.,None
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGAMTVersion1_version=1.0.0.csv ADDED
@@ -0,0 +1,610 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ GAMT (HGNC:4136),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ GAMT (HGNC:4136),PVS1,Very Strong,"Nonsense-mediated decay predicted.
9
+
10
+
11
+
12
+
13
+ CCDS VCEP notes:
14
+ Loss of function (LOF) of GAMT is a known mechanism of disease for guanidinoacetate methyltransferase deficiency (GAMT-D). There are examples of various LOF variants, including nonsense and frameshift, in GAMT in individuals with GAMT-D (
15
+ https://databases.lovd.nl/shared/variants/GAMT?search_var_status=%3D%22Marked%22%7C%3D%22Public%22
16
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
17
+
18
+
19
+ Nonsense and frameshift variants
20
+
21
+
22
+
23
+
24
+ All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 6) or the last 50 bases of the penultimate exon of the gene (exon 5, c.520). In that case, PVS1_Moderate will be applied.
25
+
26
+
27
+
28
+
29
+ Splice site variants (+1, +2, -1, -2)
30
+
31
+
32
+
33
+
34
+ All canonical splice site pairs in GAMT are GT-AG.
35
+
36
+
37
+ For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
38
+
39
+
40
+ For the predicted in frame/out of frame consequences of exon skipping and assigned strength of PVS1, see Appendix 1.
41
+
42
+
43
+ If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
44
+
45
+
46
+ To apply PVS1, splice site variants must have no detectable nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. Otherwise, the PVS1 strength should be reduced accordingly.
47
+
48
+
49
+ Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
50
+
51
+
52
+
53
+
54
+ Initiator codon variants
55
+
56
+
57
+
58
+
59
+ To our knowledge, initiator codon variants have not been reported in GAMT (01/2019) but may occur.
60
+
61
+
62
+ All initiator codon variants will meet PVS1_Moderate. The next in-frame methionine is at amino acid position 42 (based on NP_000147.1).
63
+
64
+
65
+
66
+
67
+ Deletions (single or multi exon)
68
+
69
+
70
+
71
+
72
+ If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
73
+
74
+
75
+ If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
76
+
77
+
78
+ If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
79
+
80
+
81
+ Appendix 1 can be used to predict the consequences of single exon deletions.
82
+
83
+
84
+
85
+
86
+ Duplications
87
+
88
+
89
+
90
+
91
+ Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",None
92
+ GAMT (HGNC:4136),PVS1,Strong,"In frame loss of >10% of the protein.
93
+
94
+
95
+
96
+
97
+ CCDS VCEP notes:
98
+ Loss of function (LOF) of GAMT is a known mechanism of disease for guanidinoacetate methyltransferase deficiency (GAMT-D). There are examples of various LOF variants, including nonsense and frameshift, in GAMT in individuals with GAMT-D (
99
+ https://databases.lovd.nl/shared/variants/GAMT?search_var_status=%3D%22Marked%22%7C%3D%22Public%22
100
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
101
+
102
+
103
+ Nonsense and frameshift variants
104
+
105
+
106
+
107
+
108
+ All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 6) or the last 50 bases of the penultimate exon of the gene (exon 5, c.520). In that case, PVS1_Moderate will be applied.
109
+
110
+
111
+
112
+
113
+ Splice site variants (+1, +2, -1, -2)
114
+
115
+
116
+
117
+
118
+ All canonical splice site pairs in GAMT are GT-AG.
119
+
120
+
121
+ For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
122
+
123
+
124
+ For the predicted in frame/out of frame consequences of exon skipping and assigned strength of PVS1, see Appendix 1.
125
+
126
+
127
+ If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
128
+
129
+
130
+ To apply PVS1, splice site variants must have no detectable nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. Otherwise, the PVS1 strength should be reduced accordingly.
131
+
132
+
133
+ Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
134
+
135
+
136
+
137
+
138
+ Initiator codon variants
139
+
140
+
141
+
142
+
143
+ To our knowledge, initiator codon variants have not been reported in GAMT (01/2019) but may occur.
144
+
145
+
146
+ All initiator codon variants will meet PVS1_Moderate. The next in-frame methionine is at amino acid position 42 (based on NP_000147.1).
147
+
148
+
149
+
150
+
151
+ Deletions (single or multi exon)
152
+
153
+
154
+
155
+
156
+ If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
157
+
158
+
159
+ If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
160
+
161
+
162
+ If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
163
+
164
+
165
+ Appendix 1 can be used to predict the consequences of single exon deletions.
166
+
167
+
168
+
169
+
170
+ Duplications
171
+
172
+
173
+
174
+
175
+ Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",Strength
176
+ GAMT (HGNC:4136),PVS1,Moderate,"Initiator codon variant.
177
+
178
+
179
+ Single exon or larger deletion resulting in loss of <10% of the protein.
180
+
181
+
182
+
183
+
184
+ CCDS VCEP notes:
185
+ Loss of function (LOF) of GAMT is a known mechanism of disease for guanidinoacetate methyltransferase deficiency (GAMT-D). There are examples of various LOF variants, including nonsense and frameshift, in GAMT in individuals with GAMT-D (
186
+ https://databases.lovd.nl/shared/variants/GAMT?search_var_status=%3D%22Marked%22%7C%3D%22Public%22
187
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
188
+
189
+
190
+ Nonsense and frameshift variants
191
+
192
+
193
+
194
+
195
+ All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 6) or the last 50 bases of the penultimate exon of the gene (exon 5, c.520). In that case, PVS1_Moderate will be applied.
196
+
197
+
198
+
199
+
200
+ Splice site variants (+1, +2, -1, -2)
201
+
202
+
203
+
204
+
205
+ All canonical splice site pairs in GAMT are GT-AG.
206
+
207
+
208
+ For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
209
+
210
+
211
+ For the predicted in frame/out of frame consequences of exon skipping and assigned strength of PVS1, see Appendix 1.
212
+
213
+
214
+ If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
215
+
216
+
217
+ To apply PVS1, splice site variants must have no detectable nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. Otherwise, the PVS1 strength should be reduced accordingly.
218
+
219
+
220
+ Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
221
+
222
+
223
+
224
+
225
+ Initiator codon variants
226
+
227
+
228
+
229
+
230
+ To our knowledge, initiator codon variants have not been reported in GAMT (01/2019) but may occur.
231
+
232
+
233
+ All initiator codon variants will meet PVS1_Moderate. The next in-frame methionine is at amino acid position 42 (based on NP_000147.1).
234
+
235
+
236
+
237
+
238
+ Deletions (single or multi exon)
239
+
240
+
241
+
242
+
243
+ If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
244
+
245
+
246
+ If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
247
+
248
+
249
+ If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
250
+
251
+
252
+ Appendix 1 can be used to predict the consequences of single exon deletions.
253
+
254
+
255
+
256
+
257
+ Duplications
258
+
259
+
260
+
261
+
262
+ Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",Strength
263
+ GAMT (HGNC:4136),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
264
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
265
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
266
+ GAMT (HGNC:4136),PS1,Strong,"CCDS VCEP notes:
267
+
268
+
269
+
270
+
271
+ This criterion is applicable as described.
272
+
273
+
274
+ If the variant is in the last 3 nucleotides of an exon, further analysis using splicing site prediction algorithms (see PP3) and data from the literature (if available) is required to investigate the impact on splicing.",None
275
+ GAMT (HGNC:4136),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
276
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
277
+ GAMT (HGNC:4136),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
278
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
279
+ GAMT (HGNC:4136),PS3,Strong,"RT-PCR evidence of mis-splicing for non-canonical +1, +2, -2, -1 intronic variants with no evidence of normal splice products.
280
+ GAMT specifications:
281
+ Any variant meeting the description for splicing assays below can meet PS3 or PS3_Supporting, and variants meeting the description for in vitro activity assays can meet PS3_Supporting. If a variant meets the description for both e.g., a splice site variant with evidence of abnormal splicing and deficient GAMT activity in vitro, PS3 must only be counted once.
282
+ In vitro expression
283
+ PS3_Supporting can be assigned if a variant is expressed in GAMT-deficient fibroblasts or HeLa cells and has <15% of the control value, for values published in Mercimek-Mahmutoglu et al, 2014, PMID 24415674; Mercimek-Mahmutoglu et al, 2016, PMID 26319512; or DesRoches et al, 2016, PMID 26003046. See Appendix 2 for further details on GAMT functional assays.
284
+ Splicing assays
285
+ For non-canonical splicing variants, use PS3 if there is RT-PCR and/or RNA sequencing evidence demonstrating only abnormal splice products, with no evidence of normal splicing.
286
+
287
+
288
+ Evidence of abnormal splicing includes transcripts of alternative length or with specific intron or exon inclusion/exclusion. These studies can be performed on mRNA extracted from patient-derived cells, or by inserting the mutant genomic DNA into plasmid vectors and introducing these into human or other mammalian host cells.
289
+
290
+
291
+ Note that in patients who are compound heterozygotes for a splicing variant and another variants type that does not disrupt splicing, such as a missense variant, evidence of normal splicing is expected. However, the presence of normal splice products could complicate the assessment of the impact of the splice variant. Therefore, if there is any evidence of normal splice products, either when using RNA from patient cells or in an in vitro expression system, use PS3_Supporting.
292
+
293
+
294
+ PP3 may also be used for non-canonical splice variants meeting PS3 or PS3_Supporting.",Disease-specific
295
+ GAMT (HGNC:4136),PS3,Supporting,"<15% control activity when variant is expressed in GAMT-deficient fibroblasts or HeLa cells, as reported in PMID 24415674, 26003046, and 26319512.
296
+
297
+
298
+ RT-PCR evidence of mis-splicing for non-canonical intronic variants with evidence of normal splice products.
299
+ GAMT specifications:
300
+ Any variant meeting the description for splicing assays below can meet PS3 or PS3_Supporting, and variants meeting the description for in vitro activity assays can meet PS3_Supporting. If a variant meets the description for both e.g., a splice site variant with evidence of abnormal splicing and deficient GAMT activity in vitro, PS3 must only be counted once.
301
+ In vitro expression
302
+ PS3_Supporting can be assigned if a variant is expressed in GAMT-deficient fibroblasts or HeLa cells and has <15% of the control value, for values published in Mercimek-Mahmutoglu et al, 2014, PMID 24415674; Mercimek-Mahmutoglu et al, 2016, PMID 26319512; or DesRoches et al, 2016, PMID 26003046. See Appendix 2 for further details on GAMT functional assays.
303
+ Splicing assays
304
+ For non-canonical splicing variants, use PS3 if there is RT-PCR and/or RNA sequencing evidence demonstrating only abnormal splice products, with no evidence of normal splicing.
305
+
306
+
307
+ Evidence of abnormal splicing includes transcripts of alternative length or with specific intron or exon inclusion/exclusion. These studies can be performed on mRNA extracted from patient-derived cells, or by inserting the mutant genomic DNA into plasmid vectors and introducing these into human or other mammalian host cells.
308
+
309
+
310
+ Note that in patients who are compound heterozygotes for a splicing variant and another variants type that does not disrupt splicing, such as a missense variant, evidence of normal splicing is expected. However, the presence of normal splice products could complicate the assessment of the impact of the splice variant. Therefore, if there is any evidence of normal splice products, either when using RNA from patient cells or in an in vitro expression system, use PS3_Supporting.
311
+
312
+
313
+ PP3 may also be used for non-canonical splice variants meeting PS3 or PS3_Supporting.",Disease-specific
314
+ GAMT (HGNC:4136),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
315
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
316
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",NA
317
+ GAMT (HGNC:4136),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
318
+ GAMT (HGNC:4136),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
319
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
320
+ GAMT (HGNC:4136),PM2,Supporting,"Allele frequency <0.0004 (<0.04%) in all populations in gnomAD.
321
+
322
+
323
+
324
+
325
+ CCDS VCEP notes:
326
+ It is acceptable for a GAMT variant to be present in controls, if heterozygous, because GAMT-D is a recessive disorder. Homozygotes should not be seen in a population database, such as gnomAD, because the penetrance of this condition in individuals with biallelic pathogenic variants is expected to be 100%.
327
+
328
+
329
+ GAMT specifications:
330
+
331
+
332
+
333
+
334
+ All subpopulations in gnomAD must have a maximum allele frequency less than 0.0004 (the highest population minor allele frequency of the most common pathogenic GAMT variant, c.327G>A, in gnomAD). Any variant with a frequency below this cutoff will meet PM2_Supporting (See Appendix 3).
335
+ Note – PM2 will NOT be used at moderate strength; PM2 will only be applied as a Supporting criterion.
336
+
337
+
338
+ If homozygotes are observed, the variant will meet BS2 (assuming 100% penetrance for an individual with 2 pathogenic variants in trans).","Strength,Disease-specific"
339
+ GAMT (HGNC:4136),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
340
+ Note: This requires testing of parents (or offspring) to determine phase.",
341
+ GAMT (HGNC:4136),PM3,Very Strong,"Consult specifications for assigning strength of evidence for PM3.
342
+
343
+
344
+ GAMT specifications:
345
+
346
+
347
+
348
+
349
+ Following SVI guidance for PM3 (
350
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
351
+ ), use the scoring system below.
352
+
353
+
354
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
355
+ GAMT (HGNC:4136),PM3,Strong,"Consult specifications for assigning strength of evidence for PM3.
356
+
357
+
358
+ GAMT specifications:
359
+
360
+
361
+
362
+
363
+ Following SVI guidance for PM3 (
364
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
365
+ ), use the scoring system below.
366
+
367
+
368
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
369
+ GAMT (HGNC:4136),PM3,Moderate,"Consult specifications for assigning strength of evidence for PM3.
370
+
371
+
372
+ GAMT specifications:
373
+
374
+
375
+
376
+
377
+ Following SVI guidance for PM3 (
378
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
379
+ ), use the scoring system below.
380
+
381
+
382
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",None
383
+ GAMT (HGNC:4136),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
384
+ GAMT (HGNC:4136),PM4,Moderate,"CCDS VCEP notes:
385
+
386
+
387
+
388
+
389
+ Stop loss variants in GAMT have not been reported, as far as we are aware.
390
+
391
+
392
+
393
+
394
+ GAMT specifications:
395
+ Use this rule “as is” for in frame deletions and insertions of 2 or more amino acids, but downgrade to PM4_Supporting for single amino acid deletions and insertions.",None
396
+ GAMT (HGNC:4136),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
397
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
398
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
399
+ GAMT (HGNC:4136),PM5,Moderate,"GAMT specifications:
400
+
401
+
402
+
403
+
404
+ If the pathogenicity of another missense change at the same amino acid residue is unknown, determine its pathogenicity using these specifications in order to determine if this criterion can be used. If the variant is pathogenic, use PM5. If the variant is likely pathogenic, use PM5_Supporting.",None
405
+ GAMT (HGNC:4136),PM5,Supporting,"GAMT specifications:
406
+
407
+
408
+
409
+
410
+ If the pathogenicity of another missense change at the same amino acid residue is unknown, determine its pathogenicity using these specifications in order to determine if this criterion can be used. If the variant is pathogenic, use PM5. If the variant is likely pathogenic, use PM5_Supporting.",Strength
411
+ GAMT (HGNC:4136),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
412
+ GAMT (HGNC:4136),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
413
+ Note: May be used as stronger evidence with increasing segregation data.",NA
414
+ GAMT (HGNC:4136),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
415
+ GAMT (HGNC:4136),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
416
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
417
+ GAMT (HGNC:4136),PP3,Moderate,"Non-canonical splice site variant predicted to be more deleterious than a previously observed pathogenic variant at the same nucleotide.
418
+
419
+
420
+
421
+
422
+ GAMT specifications:
423
+
424
+
425
+
426
+
427
+ For missense changes, those with a REVEL score more than 0.75 will meet PP3.
428
+
429
+
430
+ For in frame insertions and deletions, use PROVEAN and Mutation Taster. Results must be consistent to count.
431
+
432
+
433
+ For non-canonical splice site variants (e.g., +3, -3), use SpliceAI ((
434
+ https://spliceailookup.broadinstitute.org/
435
+ ) and varSEAK (
436
+ https://varseak.bio/
437
+ ). Results must be consistent to apply this criterion.
438
+
439
+
440
+ For SpliceAI, any donor loss or acceptor loss with a score >0.5. For varSEAK, any variant with splicing class 4 or 5. Evidence for creation of a cryptic splice site should also be assessed.
441
+
442
+
443
+ Do not apply this rule for canonical splice site changes meeting PVS1.
444
+
445
+
446
+ Upgrade to PP3_Moderate if a different non-canonical splice site variant at the same position is known to be pathogenic AND the newly observed variant is at least as deleterious as the previously observed variant, based on in silico prediction (Human Splicing Finder, MaxEntScan) e.g. variant being interpreted is +3C>G, previously reported variant is +3C>T and +3C>G is predicted to be more deleterious than +3C>T.",Strength
447
+ GAMT (HGNC:4136),PP3,Supporting,"REVEL score >0.75 for missense variants.
448
+
449
+
450
+ In frame deletion or insertion predicted deleterious by PROVEAN and MutationTaster.
451
+
452
+
453
+ Predicted impact on splicing by SpliceAI and varSEAK.
454
+
455
+
456
+
457
+
458
+ GAMT specifications:
459
+
460
+
461
+
462
+
463
+ For missense changes, those with a REVEL score more than 0.75 will meet PP3.
464
+
465
+
466
+ For in frame insertions and deletions, use PROVEAN and Mutation Taster. Results must be consistent to count.
467
+
468
+
469
+ For non-canonical splice site variants (e.g., +3, -3), use SpliceAI ((
470
+ https://spliceailookup.broadinstitute.org/
471
+ ) and varSEAK (
472
+ https://varseak.bio/
473
+ ). Results must be consistent to apply this criterion.
474
+
475
+
476
+ For SpliceAI, any donor loss or acceptor loss with a score >0.5. For varSEAK, any variant with splicing class 4 or 5. Evidence for creation of a cryptic splice site should also be assessed.
477
+
478
+
479
+ Do not apply this rule for canonical splice site changes meeting PVS1.
480
+
481
+
482
+ Upgrade to PP3_Moderate if a different non-canonical splice site variant at the same position is known to be pathogenic AND the newly observed variant is at least as deleterious as the previously observed variant, based on in silico prediction (Human Splicing Finder, MaxEntScan) e.g. variant being interpreted is +3C>G, previously reported variant is +3C>T and +3C>G is predicted to be more deleterious than +3C>T.",None
483
+ GAMT (HGNC:4136),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
484
+ GAMT (HGNC:4136),PP4,Strong,"Any combination of two or more of the data types listed for PP4_Moderate and PP4 OR significantly decreased creatine peak and visible GAA peak on 1H-MRS (see points scheme in main document)
485
+ Note: Do not apply this rule if the variant meets BA1, or otherwise meets criteria for benign or likely benign status.
486
+
487
+
488
+ GAMT specifications:
489
+ The following table shows the weight used for PP4 depending upon the data that is available. Assign points for each type of data present, and then add up the points.
490
+ PP4 = 1-2 points
491
+ PP4_Moderate = 3 points
492
+ PP4_Strong = 4 points
493
+ Two or more data types are recommended to reach moderate and required to reach strong.
494
+
495
+
496
+
497
+
498
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
499
+
500
+
501
+ For PP4 to be applied at strong, full GAMT gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
502
+ GAMT (HGNC:4136),PP4,Moderate,"Elevated GAA and low or low normal creatine in urine, OR elevated GAA and low creatine in plasma OR
503
+
504
+
505
+ Significantly decreased creatine peak on 1H-MRS (GAA not measured/not reported) OR
506
+
507
+
508
+ GAMT activity <5% normal in fibroblasts.
509
+ Note: Do not apply this rule if the variant meets BA1, or otherwise meets criteria for benign or likely benign status.
510
+
511
+
512
+
513
+
514
+ GAMT specifications:
515
+ The following table shows the weight used for PP4 depending upon the data that is available. Assign points for each type of data present, and then add up the points.
516
+ PP4 = 1-2 points
517
+ PP4_Moderate = 3 points
518
+ PP4_Strong = 4 points
519
+ Two or more data types are recommended to reach moderate and required to reach strong.
520
+
521
+
522
+
523
+
524
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
525
+
526
+
527
+ For PP4 to be applied at strong, full GAMT gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Strength
528
+ GAMT (HGNC:4136),PP4,Supporting,"Elevated GAA in plasma or urine (with no creatine measurement)
529
+ Note: Do not apply this rule if the variant meets BA1, or otherwise meets criteria for benign or likely benign status.
530
+
531
+
532
+
533
+
534
+ GAMT specifications:
535
+ The following table shows the weight used for PP4 depending upon the data that is available. Assign points for each type of data present, and then add up the points.
536
+ PP4 = 1-2 points
537
+ PP4_Moderate = 3 points
538
+ PP4_Strong = 4 points
539
+ Two or more data types are recommended to reach moderate and required to reach strong.
540
+
541
+
542
+
543
+
544
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
545
+
546
+
547
+ For PP4 to be applied at strong, full GAMT gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
548
+ GAMT (HGNC:4136),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
549
+ GAMT (HGNC:4136),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
550
+ GAMT (HGNC:4136),BA1,Stand Alone,"Allele frequency >0.003 (0.3%) in gnomAD in any continental population in gnomAD with >2000 alleles.
551
+
552
+
553
+ GAMT specifications:
554
+
555
+
556
+
557
+
558
+ Any variant with a frequency >0.003 (based on the estimated prevalence 1 in 114,000, PMID 24071436) in gnomAD (max allelic contribution = 100%; max genetic contribution = 100%)(See Appendix 3).
559
+
560
+
561
+ Use the highest population minor allele frequency (MAF) in any given continental population with >2,000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383).",Disease-specific
562
+ GAMT (HGNC:4136),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
563
+ GAMT (HGNC:4136),BS1,Strong,"Allele frequency >0.001 (0.1%) in gnomAD in any continental population in gnomAD with >2000 alleles.
564
+ GAMT specifications:
565
+
566
+
567
+
568
+
569
+ Any variant with a frequency >0.001 (based on the estimated prevalence 1 in 114,000 (PMID: 24071436) in gnomAD (max allelic contribution = 40%; max genetic contribution = 100%)(See Appendix 3).
570
+
571
+
572
+ Use the highest population minor allele frequency (MAF) in any given continental population with >2,000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383).",Disease-specific
573
+ GAMT (HGNC:4136),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
574
+ GAMT (HGNC:4136),BS2,Strong,Observed in the homozygous state in a healthy adult.,None
575
+ GAMT (HGNC:4136),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
576
+ GAMT (HGNC:4136),BS3,Supporting,"30% normal GAA activity when the variant is expressed in a heterologous cell type.
577
+ GAMT specifications:
578
+ In vitro assays in which a variant is expressed in GAMT-deficient cultured cells (e.g. GAMT-deficient fibroblasts) or in-fusion High-Fidelity cloning of GAMT transcript and site directed mutagenesis to generate missense variant overexpressed in HeLa cells and measurement of GAMT activity in cells for wild-type and missense variant. Any variant with enzyme activity at or above 30% of normal in the following publications meets BS3_Supporting (Mercimek-Mahmutoglu et al, 2014; PMID 24415674; Mercimek-Mahmutoglu et al, 2016, PMID 26319512; DesRoches et al, 2016, PMID 26003046).","Strength,Disease-specific"
579
+ GAMT (HGNC:4136),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
580
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
581
+ GAMT (HGNC:4136),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
582
+ GAMT (HGNC:4136),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
583
+ GAMT (HGNC:4136),BP2,Supporting,"Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder; or observed in cis with a pathogenic variant in any inheritance pattern.
584
+ GAMT specifications:
585
+ Observed in cis with a pathogenic variant (to take AR inheritance into account).",None
586
+ GAMT (HGNC:4136),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
587
+ GAMT (HGNC:4136),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
588
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
589
+ GAMT (HGNC:4136),BP4,Supporting,"GAMT specifications:
590
+
591
+
592
+
593
+
594
+ For missense changes, REVEL score <0.5.
595
+
596
+
597
+ For in frame insertions and deletions, use PROVEAN and Mutation Taster. Results must be consistent to count.
598
+
599
+
600
+ For non-canonical splice site variants, use SpliceAI (
601
+ https://spliceailookup.broadinstitute.org/
602
+ ) and varSEAK (
603
+ https://varseak.bio/
604
+ ) to assess the impact of variants that are not +/-1 or 2 canonical splice site variants. For SpliceAI, this criterion can be applied for scores <0.2, and for varSEAK class 1 and 2. If there is any evidence for possible creation of a cryptic splice site, this criterion should not be applied.",None
605
+ GAMT (HGNC:4136),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
606
+ GAMT (HGNC:4136),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
607
+ GAMT (HGNC:4136),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
608
+ GAMT (HGNC:4136),BP7,Supporting,"A synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.
609
+ GAMT specifications:
610
+ Apply this criterion as described.",None
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGAMTVersion2.0.0_version=2.0.0.csv ADDED
@@ -0,0 +1,406 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ GAMT (HGNC:4136),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ GAMT (HGNC:4136),PVS1,Very Strong,"Nonsense-mediated decay predicted
9
+
10
+
11
+ Nonsense and frameshift variants
12
+
13
+
14
+
15
+
16
+ All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 6) or the last 50 bases of the penultimate exon of the gene (exon 5, c.520). In that case, PVS1_Strong or PVS1_Moderate will be applied, depending on whether >10% or <10% of the protein is lost.
17
+
18
+
19
+
20
+
21
+ Splice site variants (+1, +2, -1, -2)
22
+
23
+
24
+
25
+
26
+ All canonical splice site pairs in GAMT are GT-AG.
27
+
28
+
29
+ For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
30
+
31
+
32
+ For the predicted in frame/out of frame consequences of exon skipping and assigned strength of PVS1, see Appendix 1.
33
+
34
+
35
+ If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
36
+
37
+
38
+ Use SpliceAI to look for nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. Otherwise, the PVS1 strength should be reduced accordingly.
39
+
40
+
41
+ Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
42
+
43
+
44
+
45
+
46
+ Deletions (single or multi exon)
47
+
48
+
49
+
50
+
51
+ If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. 
52
+
53
+
54
+ If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
55
+
56
+
57
+ If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
58
+
59
+
60
+ If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
61
+ Appendix 1 can be used to predict the consequences of single exon deletions.
62
+
63
+
64
+
65
+
66
+ Duplications
67
+
68
+
69
+
70
+
71
+ Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",None
72
+ GAMT (HGNC:4136),PVS1,Strong,"Single exon or larger deletion resulting in loss of >10% of the protein.
73
+  
74
+
75
+
76
+ Nonsense and frameshift variants
77
+
78
+
79
+
80
+
81
+ All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 6) or the last 50 bases of the penultimate exon of the gene (exon 5, c.520). In that case, PVS1_Strong will be applied if >10% of the protein is lost.
82
+
83
+
84
+
85
+
86
+ Splice site variants (+1, +2, -1, -2)
87
+
88
+
89
+
90
+
91
+ All canonical splice site pairs in GAMT are GT-AG.
92
+
93
+
94
+ For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
95
+
96
+
97
+ Use SpliceAI to look for nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. 
98
+
99
+
100
+ For considerations for strength at which PVS1 may be applied see Appendix 1.
101
+
102
+
103
+ If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
104
+
105
+
106
+ Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
107
+
108
+
109
+
110
+
111
+ Deletions (single or multi exon)
112
+
113
+
114
+
115
+
116
+ If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD.
117
+
118
+
119
+  If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
120
+
121
+
122
+ If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
123
+
124
+
125
+ If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
126
+ Appendix 1 can be used to predict the consequences of single exon deletions.
127
+
128
+
129
+
130
+
131
+ Duplications
132
+
133
+
134
+
135
+
136
+ Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",Strength
137
+ GAMT (HGNC:4136),PVS1,Moderate,"Single exon or larger deletion resulting in loss of <10% of the protein.
138
+    
139
+
140
+
141
+ Nonsense and frameshift variants
142
+
143
+
144
+
145
+
146
+ All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 6) or the last 50 bases of the penultimate exon of the gene (exon 5, c.520). In that case, PVS1_Moderate will be applied if <10% of the protein is lost.
147
+
148
+
149
+
150
+
151
+ Splice site variants (+1, +2, -1, -2)
152
+
153
+
154
+
155
+
156
+ All canonical splice site pairs in GAMT are GT-AG.
157
+
158
+
159
+ Use SpliceAI to look for nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. 
160
+
161
+
162
+ For any canonical splice site variant (+1, +2, -1, -2), if RT-PCR data predicts in-frame loss of <10% of the protein, apply PVS1_Moderate.
163
+
164
+
165
+ If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
166
+
167
+
168
+ Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
169
+
170
+
171
+
172
+
173
+ Initiator codon variants
174
+
175
+
176
+
177
+
178
+ All initiator codon variants will meet PVS1_Moderate. The next in-frame methionine is at amino acid position 42 (based on NP_000147.1).
179
+
180
+
181
+
182
+
183
+ Deletions (single or multi exon)
184
+
185
+
186
+
187
+
188
+ If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. 
189
+
190
+
191
+ If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
192
+
193
+
194
+ If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
195
+
196
+
197
+ If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
198
+ Appendix 1 can be used to predict the consequences of single exon deletions.
199
+
200
+
201
+
202
+
203
+ Duplications
204
+
205
+
206
+
207
+
208
+ Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",Strength
209
+ GAMT (HGNC:4136),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
210
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
211
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
212
+ GAMT (HGNC:4136),PS1,Strong,"This criterion is applicable for any variant resulting in the same amino acid change as a variant that has been previously established as pathogenic by the CCDS VCEP, by assessment using these criteria, regardless of nucleotide change.
213
+
214
+
215
+ If the variant is in the last 3 nucleotides of an exon, further analysis using splicing site prediction algorithms (see PP3) and data from the literature (if available) is required to investigate the impact on splicing.
216
+
217
+
218
+ PS1 may also be applied for splicing variants under specific circumstances (see Table 3 in PMID: 37352859).",General recommendation
219
+ GAMT (HGNC:4136),PS1,Moderate,"This criterion is applicable for any variant resulting in the same amino acid change as a variant that has been previously established as likely pathogenic by the CCDS VCEP, by assessment using these criteria, regardless of nucleotide change.
220
+
221
+
222
+ If the variant is in the last 3 nucleotides of an exon, further analysis using splicing site prediction algorithms (see PP3) and data from the literature (if available) is required to investigate the impact on splicing.
223
+
224
+
225
+ PS1_Moderate may also be applied for splicing variants under specific circumstances (see Table 3 in PMID: 37352859).","General recommendation,Strength"
226
+ GAMT (HGNC:4136),PS1,Supporting,PS1_Supporting may be applied for splicing variants under specific circumstances (see Table 3 in PMID: 37352859).,"General recommendation,Strength"
227
+ GAMT (HGNC:4136),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
228
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
229
+ GAMT (HGNC:4136),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
230
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
231
+ GAMT (HGNC:4136),PS3,Supporting,"PS3_Supporting can be assigned if a variant is expressed in GAMT-deficient fibroblasts, HeLa cells and has <15% of the control value, for values published in Mercimek-Mahmutoglu et al, 2014, PMID 24415674; Mercimek-Mahmutoglu et al, 2016, PMID 26319512; or DesRoches et al, 2016, PMID 26003046.","Disease-specific,Strength"
232
+ GAMT (HGNC:4136),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
233
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
234
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",NA
235
+ GAMT (HGNC:4136),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
236
+ GAMT (HGNC:4136),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
237
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
238
+ GAMT (HGNC:4136),PM2,Supporting,"Allele frequency <0.0004 (<0.04%) in all populations in gnomAD.
239
+
240
+
241
+ It is acceptable for a GAMT variant to be present in controls, if heterozygous, because GAMT-D is a recessive disorder. Homozygotes should not be seen in a population database, such as gnomAD, because the penetrance of this condition in individuals with biallelic pathogenic variants is expected to be 100% and the condition is expected to present with severe symptoms early in life.
242
+
243
+
244
+ GAMT specifications:
245
+
246
+
247
+
248
+
249
+ All subpopulations in gnomAD v4.0 must have a maximum allele frequency less than 0.0004 (the highest population minor allele frequency of the most common pathogenic GAMT variant, c.327G>A, in gnomAD). Any variant with a frequency below this cutoff will meet PM2_Supporting.
250
+
251
+
252
+ If homozygotes are observed, or variant is confirmed in trans with a known pathogenic variant, the variant will meet BS2 (assuming 100% penetrance for an individual with 2 pathogenic variants in trans).","Disease-specific,Strength"
253
+ GAMT (HGNC:4136),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
254
+ Note: This requires testing of parents (or offspring) to determine phase.",
255
+ GAMT (HGNC:4136),PM3,Very Strong,"Follow SVI guidance for PM3 (
256
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
257
+ ).
258
+
259
+
260
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
261
+ GAMT (HGNC:4136),PM3,Strong,"Follow SVI guidance for PM3 (
262
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
263
+ ).
264
+
265
+
266
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
267
+ GAMT (HGNC:4136),PM3,Moderate,"Follow SVI guidance for PM3 (
268
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
269
+ ).
270
+
271
+
272
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",None
273
+ GAMT (HGNC:4136),PM3,Supporting,"Follow SVI guidance for PM3 (
274
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
275
+ ).
276
+
277
+
278
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
279
+ GAMT (HGNC:4136),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
280
+ GAMT (HGNC:4136),PM4,Moderate,"GAMT specifications: Use this rule “as is” for stop loss variant, and in frame deletions and insertions of 2 or more amino acids, but downgrade to PM4_Supporting for single amino acid deletions and insertions.",None
281
+ GAMT (HGNC:4136),PM4,Supporting,Downgrade to PM4_Supporting for single amino acid deletions and insertions.,Strength
282
+ GAMT (HGNC:4136),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
283
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
284
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
285
+ GAMT (HGNC:4136),PM5,Moderate,"This criterion is applicable for any variant resulting in a different amino acid change, at the same amino acid position, as a variant that has been previously established as pathogenic by the CCDS VCEP, by assessment using these criteria, regardless of nucleotide change.
286
+
287
+
288
+ If the variant is in the last 3 nucleotides of an exon, further analysis using splicing site prediction algorithms (see PP3) and data from the literature (if available) is required to investigate the impact on splicing. 
289
+
290
+
291
+ If the variant is likely pathogenic, use PM5_Supporting.",None
292
+ GAMT (HGNC:4136),PM5,Supporting,"This criterion is applicable for any variant resulting in a different amino acid change, at the same amino acid position, as a variant that has been previously established as likely pathogenic by the CCDS VCEP, by assessment using these criteria, regardless of nucleotide change.
293
+
294
+
295
+ If the variant is in the last 3 nucleotides of an exon, further analysis using splicing site prediction algorithms (see PP3) and data from the literature (if available) is required to investigate the impact on splicing.  
296
+
297
+
298
+ If the variant is likely pathogenic, use PM5.",Strength
299
+ GAMT (HGNC:4136),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
300
+ GAMT (HGNC:4136),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
301
+ Note: May be used as stronger evidence with increasing segregation data.",NA
302
+ GAMT (HGNC:4136),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
303
+ GAMT (HGNC:4136),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
304
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
305
+ GAMT (HGNC:4136),PP3,Strong,"Missense variant with a REVEL score equal to or >0.932 (based on guidance from Pejaver et al, 2022, PMID: 36413997).","General recommendation,Strength"
306
+ GAMT (HGNC:4136),PP3,Moderate,"Missense variant with a REVEL score 0.773-0.932 (based on guidance from Pejaver et al, 2022, PMID: 36413997).","General recommendation,Strength"
307
+ GAMT (HGNC:4136),PP3,Supporting,"Missense variant with a REVEL score 0.644-0.773 (based on guidance from Pejaver et al, 2022, PMID: 36413997).
308
+
309
+
310
+ In frame deletion or insertion predicted deleterious by PROVEAN and MutationTaster. Results must be consistent to count.
311
+
312
+
313
+ For non-canonical splice site variants (e.g., +3, -3), predict the impact on splicing by using SpliceAI,
314
+ https://spliceailookup.broadinstitute.org/
315
+ , and apply PP3 for a score equal to or >0.2 (as indicated in PMID: 37352859, Table 1 and Figure 4). Assess the possibility of activation of cryptic splice sites.","General recommendation,None"
316
+ GAMT (HGNC:4136),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
317
+ GAMT (HGNC:4136),PP4,Strong,"4 points based on any combination of the following. Two or more data types are required to apply PP4_Strong:
318
+
319
+
320
+
321
+
322
+ Elevated urine guanidinoacetate with or without low or low normal creatine (1 point).
323
+
324
+
325
+ Elevated plasma guanidinoacetate with or without low or low normal creatine (2 points).
326
+
327
+
328
+ Significantly decreased creatine peak in brain magnetic resonance spectroscopy with or without visible guanidinoacetate peak (3 points).
329
+
330
+
331
+ GAMT enzyme activity <5% of normal (3 points).
332
+
333
+
334
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
335
+
336
+
337
+ For PP4 to be applied at strong, full GAMT gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
338
+ GAMT (HGNC:4136),PP4,Moderate,"3 points based on any combination of the following. Two or more data types are recommended to apply PP4_Moderate:
339
+
340
+
341
+
342
+
343
+ Elevated urine guanidinoacetate with or without low or low normal creatine (1 point).
344
+
345
+
346
+ Elevated plasma guanidinoacetate with or without low or low normal creatine (2 points).
347
+
348
+
349
+ Significantly decreased creatine peak in brain magnetic resonance spectroscopy with or without visible guanidinoacetate peak (3 points).
350
+
351
+
352
+ GAMT enzyme activity <5% of normal (3 points).
353
+
354
+
355
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.",Strength
356
+ GAMT (HGNC:4136),PP4,Supporting,"1-2 points based on: 
357
+
358
+
359
+
360
+
361
+ Elevated urine guanidinoacetate with or without low or low normal creatine (1 point).
362
+
363
+
364
+ Elevated plasma guanidinoacetate with or without low or low normal creatine (2 points).
365
+
366
+
367
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.",Disease-specific
368
+ GAMT (HGNC:4136),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
369
+ GAMT (HGNC:4136),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
370
+ GAMT (HGNC:4136),BA1,Stand Alone,"Allele frequency >0.003 (0.3%) in gnomAD v4.0 in any continental population with >2000 alleles (based on the estimated prevalence 1 in 114,000, PMID 24071436) in gnomAD v4.0 (max allelic contribution = 100%; max genetic contribution = 100%).
371
+
372
+
373
+ Use the highest population minor allele frequency (MAF) in any given continental population with >2,000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383).",Disease-specific
374
+ GAMT (HGNC:4136),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
375
+ GAMT (HGNC:4136),BS1,Strong,"Allele frequency >0.001 (0.1%) in gnomAD v4.0 in any continental population with >2000 alleles (based on the estimated prevalence 1 in 114,000 (PMID: 24071436) in gnomAD v4.0 (max allelic contribution = 40%; max genetic contribution = 100%).
376
+
377
+
378
+ Use the highest population minor allele frequency (MAF) in any given continental population with >2,000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383).",Disease-specific
379
+ GAMT (HGNC:4136),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
380
+ GAMT (HGNC:4136),BS2,Strong,"Observed in the homozygous state in a healthy adult, or confirmed in trans with a variant that has been classified as pathogenic by the CCDS VCEP using these criteria.",None
381
+ GAMT (HGNC:4136),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
382
+ GAMT (HGNC:4136),BS3,Supporting,"In vitro assays in which a variant is expressed in GAMT-deficient cultured cells (e.g. GAMT-deficient fibroblasts) or in-fusion High-Fidelity cloning of GAMT transcript and site directed mutagenesis to generate missense variant overexpressed in HeLa cells and measurement of GAMT activity in cells for wild-type and missense variant. Any variant with enzyme activity at or above 30% of normal in the following publications meets BS3_Supporting (Mercimek-Mahmutoglu et al, 2014; PMID 24415674; Mercimek-Mahmutoglu et al, 2016, PMID 26319512; DesRoches et al, 2016, PMID 26003046).","Disease-specific,Strength"
383
+ GAMT (HGNC:4136),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
384
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
385
+ GAMT (HGNC:4136),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
386
+ GAMT (HGNC:4136),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
387
+ GAMT (HGNC:4136),BP2,Supporting,Observed in cis with a pathogenic variant (to take AR inheritance into account).,None
388
+ GAMT (HGNC:4136),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
389
+ GAMT (HGNC:4136),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
390
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
391
+ GAMT (HGNC:4136),BP4,Supporting,"REVEL score <0.29 for missense variants (based on guidance from Pejaver et al, 2022, PMID: 36413997).
392
+
393
+
394
+ In frame deletion or insertion predicted benign by PROVEAN and MutationTaster.
395
+
396
+
397
+ No predicted impact on splicing by SpliceAI,
398
+ https://spliceailookup.broadinstitute.org/
399
+ , based on a score <0.1 (as indicated in PMID: 37352859, Table 1 and Figure 4).","General recommendation,None"
400
+ GAMT (HGNC:4136),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
401
+ GAMT (HGNC:4136),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
402
+ GAMT (HGNC:4136),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
403
+ GAMT (HGNC:4136),BP7,Strong,"Experimental evidence, such as RT-PCR, shows no impact on splicing. Follow the decision tree outlined in Figure 5, Walker et al, 2023, PMID: 37352859. Note that splicing may appear normal in compound heterozygous patients if the splicing defect generates a transcript that is degraded by nonsense-mediated decay. Therefore, caution must be used when assessing the data prior to applying this code.","General recommendation,Strength"
404
+ GAMT (HGNC:4136),BP7,Supporting,"A synonymous (silent) variant OR an intronic variant at or beyond positions +7 and -21, for which SpliceAI,
405
+ https://spliceailookup.broadinstitute.org/
406
+ , predicts no impact on splicing (score <0.1).",None
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGATMVersion1.1.0_version=1.1.0.csv ADDED
@@ -0,0 +1,337 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ GATM (HGNC:4175),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ GATM (HGNC:4175),PVS1,Very Strong,"Nonsense-mediated decay predicted.
9
+
10
+
11
+ CCDS VCEP notes:
12
+ Loss of function (LOF) of GATM is a known mechanism of disease for arginine:glycine amidinotransferase deficiency (AGAT-D). There are examples of various LOF variants, including nonsense and frameshift, in GATM in individuals with AGAT-D (
13
+ https://databases.lovd.nl/shared/variants/GATM/unique
14
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
15
+
16
+
17
+ GATM specifications:
18
+ Nonsense and frameshift variants
19
+ * All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 9) or the last 50 bases of the penultimate exon (exon 8, 3’ of c.1109). In that case, PVS1_Strong or PVS1_Moderate will be applied depending on whether >10% or <10% of the protein is lost.
20
+
21
+
22
+ Splice site variants (+1, +2, -1, -2)
23
+ * All canonical splice site pairs in GATM are GT-AG.
24
+ * For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
25
+ * For the predicted in frame/out of frame consequences of exon skipping and considerations for strength of PVS1, see
26
+ Appendix 1
27
+ .
28
+ * If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
29
+ * To apply PVS1, splice site variants must have no detectable nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. Otherwise, the PVS1 strength should be reduced accordingly.
30
+ * Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
31
+
32
+
33
+ Deletions (single or multi exon)
34
+ * If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
35
+ * If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
36
+ * If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
37
+ *
38
+ Appendix 1
39
+ can be used to predict the consequences of single exon deletions.
40
+
41
+
42
+ Duplications
43
+ * Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",None
44
+ GATM (HGNC:4175),PVS1,Strong,"In frame loss of >10% of the protein.
45
+ CCDS VCEP notes:
46
+ Loss of function (LOF) of GATM is a known mechanism of disease for arginine:glycine amidinotransferase deficiency (AGAT-D). There are examples of various LOF variants, including nonsense and frameshift, in GATM in individuals with AGAT-D (
47
+ https://databases.lovd.nl/shared/variants/GATM/unique
48
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
49
+
50
+
51
+ GATM specifications:
52
+ Nonsense and frameshift variants
53
+ * All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 9) or the last 50 bases of the penultimate exon (exon 8, 3’ of c.1109). In that case, PVS1_Strong or PVS1_Moderate will be applied depending on whether >10% or <10% of the protein is lost.
54
+
55
+
56
+ Splice site variants (+1, +2, -1, -2)
57
+ * All canonical splice site pairs in GATM are GT-AG.
58
+ * For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
59
+ *Use SpliceAI and varSEAK to look for nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. 
60
+ * For considerations for strength at which PVS1 may be applied see
61
+ Appendix 1
62
+ .
63
+ * If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
64
+ * Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
65
+
66
+
67
+ Deletions (single or multi exon)
68
+ * If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
69
+ * If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
70
+ * If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
71
+ *
72
+ Appendix 1
73
+ can be used to predict the consequences of single exon deletions.
74
+
75
+
76
+ Duplications
77
+ * Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",Strength
78
+ GATM (HGNC:4175),PVS1,Moderate,"Single exon or larger deletion resulting in loss of <10% of the protein, and initiator codon variants.
79
+
80
+
81
+ CCDS VCEP notes:
82
+ Loss of function (LOF) of GATM is a known mechanism of disease for arginine:glycine amidinotransferase deficiency (AGAT-D). There are examples of various LOF variants, including nonsense and frameshift, in GATM in individuals with AGAT-D (
83
+ https://databases.lovd.nl/shared/variants/GATM/unique
84
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).  
85
+
86
+
87
+ GATM specifications:
88
+
89
+
90
+ Nonsense and frameshift variants
91
+ * All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 9) or the last 50 bases of the penultimate exon (exon 8, 3’ of c.1109). In that case, PVS1_Moderate will be applied.
92
+
93
+
94
+ Splice site variants (+1, +2, -1, -2)
95
+ * All canonical splice site pairs in GATM are GT-AG.
96
+ * For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
97
+ *Use SpliceAI and varSEAK to look for nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. 
98
+ * For considerations for strength at which PVS1 may be applied see
99
+ Appendix 1
100
+ .
101
+ * If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
102
+ * Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
103
+
104
+
105
+ Initiator codon variants
106
+ * To our knowledge, initiator codon variants have not been reported in GATM (01/2019) but may occur.
107
+ * All initiator codon variants will meet PVS1_Moderate. The next in-frame methionine is at amino acid position 130 (based on NP_001473).
108
+
109
+
110
+ Deletions (single or multi exon)
111
+ * If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
112
+ * If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
113
+ * If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
114
+ *
115
+ Appendix 1
116
+ can be used to predict the consequences of single exon deletions.
117
+
118
+
119
+ Duplications
120
+ * Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",Strength
121
+ GATM (HGNC:4175),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
122
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
123
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
124
+ GATM (HGNC:4175),PS1,Strong,"This criterion is applicable as described.
125
+
126
+
127
+ If the variant is in the last 3 nucleotides of an exon, further analysis using splicing site prediction algorithms (see PP3) and data from the literature (if available) is required to investigate the impact on splicing.",None
128
+ GATM (HGNC:4175),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
129
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
130
+ GATM (HGNC:4175),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
131
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
132
+ GATM (HGNC:4175),PS3,Strong,"RT-PCR evidence of mis-splicing for non-canonical intronic variants with no evidence of normal splice products.
133
+
134
+
135
+ GATM specifications:
136
+ Any variant meeting the description for either in vitro expression or splicing assays can meet PS3 at the strengths given. If a variant meets the description for both e.g., a splice site variant with no evidence of abnormal splicing and deficient AGAT activity in vitro, PS3 must only be applied once.
137
+
138
+
139
+ Splicing assays
140
+ For non-canonical splicing variants, use PS3 if there is RT-PCR and/or RNA sequencing evidence demonstrating only abnormal splice products, with no evidence of normal splicing.
141
+ * Evidence of abnormal splicing includes transcripts of alternative length or with specific intron or exon inclusion/exclusion. These studies can be performed on mRNA extracted from patient-derived cells, or by inserting the mutant genomic DNA into plasmid vectors and introducing these into human or other mammalian host cells.
142
+ * Note that in patients who are compound heterozygotes for a splicing variant and another variants type that does not disrupt splicing, such as a missense variant, evidence of normal splicing is expected. However, the presence of normal splice products could complicate the assessment of the impact of the splice variant. Therefore, if there is any evidence of normal splice products, either when using RNA from patient cells or in an in vitro expression system, use PS3_Supporting.
143
+ * PP3 may also be used for non-canonical splice variants meeting PS3 or PS3_Supporting.",Disease-specific
144
+ GATM (HGNC:4175),PS3,Supporting,"<15% control activity when variant is expressed in HeLa cells, as reported in PMID 27233232.
145
+ RT-PCR evidence of mis-splicing for non-canonical intronic variants with evidence of normal splice products.
146
+
147
+
148
+ GATM specifications:
149
+ Any variant meeting the description for either in vitro expression or splicing assays can meet PS3 at the strengths given. If a variant meets the description for both e.g., a splice site variant with no evidence of abnormal splicing and deficient AGAT activity in vitro, PS3 must only be applied once.  
150
+
151
+
152
+ In vitro expression
153
+ AGAT activity data from an in vitro assay in which GATM variants were overexpressed in HeLa cells has been published (DesRoches et al, 2016; PMID 27233232). Any variant with AGAT activity at or below 15% of normal in this paper meets PS3_Supporting (see
154
+ Appendix 2
155
+ for further details on AGAT functional assays).  
156
+
157
+
158
+ Splicing assays
159
+ For non-canonical splicing variants, use PS3 if there is RT-PCR and/or RNA sequencing evidence demonstrating only abnormal splice products, with no evidence of normal splicing.
160
+ * Evidence of abnormal splicing includes transcripts of alternative length or with specific intron or exon inclusion/exclusion. These studies can be performed on mRNA extracted from patient-derived cells, or by inserting the mutant genomic DNA into plasmid vectors and introducing these into human or other mammalian host cells.
161
+ * Note that in patients who are compound heterozygotes for a splicing variant and another variants type that does not disrupt splicing, such as a missense variant, evidence of normal splicing is expected. However, the presence of normal splice products could complicate the assessment of the impact of the splice variant. Therefore, if there is any evidence of normal splice products, either when using RNA from patient cells or in an in vitro expression system, use PS3_Supporting.
162
+ * PP3 may also be used for non-canonical splice variants meeting PS3 or PS3_Supporting.",Disease-specific
163
+ GATM (HGNC:4175),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
164
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
165
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",NA
166
+ GATM (HGNC:4175),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
167
+ GATM (HGNC:4175),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
168
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
169
+ GATM (HGNC:4175),PM2,Supporting,"Allele frequency <0.000055 (<0.0055%) in all populations in gnomAD.
170
+
171
+
172
+ CCDS VCEP notes: It is acceptable for a GATM variant to be present in controls, if heterozygous, because AGAT-D is a recessive disorder. Homozygotes should not be seen in a population database, such as gnomAD, because the penetrance of this condition in individuals with biallelic pathogenic variants is expected to be 100%. 
173
+
174
+
175
+ GATM specifications:
176
+
177
+
178
+
179
+
180
+ All subpopulations in gnomAD must have a maximum allele frequency less than 0.000055 (based on the prevalence of the most common suspected pathogenic variants, c.484+1G>T and p.Arg169Ter) (see Appendix 3). Note – PM2 will NOT be used at moderate strength; PM2 will only be applied as a Supporting criterion.
181
+
182
+
183
+ If homozygotes are observed, the variant will meet BS2 (assuming 100% penetrance for an individual with 2 pathogenic variants in trans).","Disease-specific,Strength"
184
+ GATM (HGNC:4175),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
185
+ Note: This requires testing of parents (or offspring) to determine phase.",
186
+ GATM (HGNC:4175),PM3,Very Strong,"Follow SVI guidance for PM3 (
187
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
188
+ ).
189
+
190
+
191
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
192
+ GATM (HGNC:4175),PM3,Strong,"Follow SVI guidance for PM3 (
193
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
194
+ ).
195
+
196
+
197
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
198
+ GATM (HGNC:4175),PM3,Moderate,"Follow SVI guidance for PM3 (
199
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
200
+ ).
201
+
202
+
203
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",None
204
+ GATM (HGNC:4175),PM3,Supporting,"Follow SVI guidance for PM3 (
205
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
206
+ ).
207
+
208
+
209
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
210
+ GATM (HGNC:4175),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
211
+ GATM (HGNC:4175),PM4,Moderate,"CCDS VCEP notes: Stop loss variants in GATM have not been reported, as far as we are aware. 
212
+
213
+
214
+ GATM specifications: Use this rule “as is” for in frame deletions and insertions of 2 or more amino acids, but downgrade to PM4_Supporting for single amino acid deletions and insertions.",None
215
+ GATM (HGNC:4175),PM4,Supporting,Downgrade to PM4_Supporting for in frame deletion/insertion of a single amino acid.,Strength
216
+ GATM (HGNC:4175),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
217
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
218
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
219
+ GATM (HGNC:4175),PM5,Moderate,"If the pathogenicity of another missense change at the same amino acid residue is unknown, determine its pathogenicity using these specifications in order to determine if this criterion can be used. If the variant is pathogenic, use PM5. If the variant is likely pathogenic, use PM5_Supporting.",None
220
+ GATM (HGNC:4175),PM5,Supporting,Missense change at an amino acid residue where a different missense change determined to be likely pathogenic has been seen before.,Strength
221
+ GATM (HGNC:4175),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
222
+ GATM (HGNC:4175),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
223
+ Note: May be used as stronger evidence with increasing segregation data.",NA
224
+ GATM (HGNC:4175),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
225
+ GATM (HGNC:4175),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
226
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
227
+ GATM (HGNC:4175),PP3,Moderate,"Non-canonical splice site variant predicted to be more deleterious, by SpliceAI and varSEAK, than a previously observed pathogenic variant at the same nucleotide.",None
228
+ GATM (HGNC:4175),PP3,Supporting,"REVEL score >0.75 for missense variants.
229
+
230
+
231
+ In frame deletion or insertion predicted deleterious by PROVEAN and MutationTaster.
232
+
233
+
234
+ Predicted impact on splicing by SpliceAI and varSEAK. GATM specifications:
235
+
236
+
237
+ For missense changes, those with a REVEL score more than 0.75 will meet PP3.
238
+
239
+
240
+ For in frame insertions and deletions, use PROVEAN and Mutation Taster. Results must be consistent to count.
241
+
242
+
243
+ For non-canonical splice site variants (e.g., +3, -3), use SpliceAI ((
244
+ https://spliceailookup.broadinstitute.org/
245
+ ) and varSEAK (
246
+ https://varseak.bio/
247
+ ). Results must be consistent to apply this criterion.
248
+
249
+
250
+ For SpliceAI, any donor loss or acceptor loss with a score >0.5. For varSEAK, any variant with splicing class 4 or 5. Evidence for creation of a cryptic splice site should also be assessed.
251
+
252
+
253
+ Do not apply this rule for canonical splice site changes meeting PVS1.",None
254
+ GATM (HGNC:4175),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
255
+ GATM (HGNC:4175),PP4,Strong,"4 or more points based on any combination of the following. Two or more data types are required to meet Strong:
256
+ • Low urine guanidinoacetate with or without low or low normal creatine (1 point)
257
+ • Low plasma guanidinoacetate with or without low or low normal creatine (2 points)
258
+ • Significantly decreased creatine peak in brain magnetic resonance spectroscopy (3 points)
259
+ • AGAT enzyme activity <5% of normal (3 points)
260
+
261
+
262
+ * Variant must meet PM2_Supporting for PP4 to apply at any strength.
263
+ * For PP4 to be applied at strong, full GATM gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
264
+ GATM (HGNC:4175),PP4,Moderate,"3 points based on any combination of the following. Two or more data types are recommended to reach moderate:
265
+ • Low urine guanidinoacetate with or without low or low normal creatine (1 point)
266
+ • Low plasma guanidinoacetate with or without low or low normal creatine (2 points)
267
+ • Significantly decreased creatine peak in brain magnetic resonance spectroscopy (3 points)
268
+ • AGAT enzyme activity <5% of normal (3 points)
269
+
270
+
271
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.",Strength
272
+ GATM (HGNC:4175),PP4,Supporting,"1-2 points based on:
273
+ • Low urine guanidinoacetate with or without low or low normal creatine (1 point)
274
+ • Low plasma guanidinoacetate with or without low or low normal creatine (2 points)
275
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.",Disease-specific
276
+ GATM (HGNC:4175),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
277
+ GATM (HGNC:4175),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
278
+ GATM (HGNC:4175),BA1,Stand Alone,"Allele frequency >0.0005 (0.05%) in gnomAD in any continental population in gnomAD with >2000 alleles.
279
+
280
+
281
+
282
+
283
+ Any variant with a frequency >0.0005 (max allelic contribution = 100% and max genetic contribution = 100% based on estimated prevalence of 1 in 3,450,000 (PMID 27233232), and penetrance of 100%) in a continental population with >2000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383).
284
+
285
+
286
+ Use the highest population minor allele frequency (MAF) in any given continental population with >2,000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383).",Disease-specific
287
+ GATM (HGNC:4175),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
288
+ GATM (HGNC:4175),BS1,Strong,"Allele frequency >0.0001 (0.01%) in gnomAD in any continental population in gnomAD with >2000 alleles.
289
+
290
+
291
+
292
+
293
+ Any variant with a frequency >0.0001 (max allelic contribution = 25% and max genetic contribution = 100% based on estimated prevalence of 1 in 3,450,000, (PMID 27233232), and penetrance of 100%) in a continental population with >2000 (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383) (see Appendix 3).
294
+
295
+
296
+ Use the highest population minor allele frequency (MAF) in any given continental population with >2,000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383).",Disease-specific
297
+ GATM (HGNC:4175),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
298
+ GATM (HGNC:4175),BS2,Strong,Observed in the homozygous state in a healthy adult.,Disease-specific
299
+ GATM (HGNC:4175),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
300
+ GATM (HGNC:4175),BS3,Supporting,">30% normal GAA activity when the variant is expressed in a heterologous cell type.
301
+ GATM specifications:
302
+ In vitro assays in which a variant is expressed in AGAT-deficient cultured cells (e.g. AGAT-deficient fibroblasts) or in-fusion High-Fidelity cloning of GATM transcript and site directed mutagenesis to generate missense variant overexpressed in HeLa cells and measurement of AGAT activity in cells for wild-type and missense variant. Any variant with enzyme activity at or above 30% of normal in DesRoches et al, 2016, PMID 27233232, meets BS3_Supporting.","Disease-specific,Strength"
303
+ GATM (HGNC:4175),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
304
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
305
+ GATM (HGNC:4175),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
306
+ GATM (HGNC:4175),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
307
+ GATM (HGNC:4175),BP2,Supporting,"Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder; or observed in cis with a pathogenic variant in any inheritance pattern.
308
+ GATM specifications:
309
+ Observed in cis with a pathogenic variant (to take AR inheritance into account).",None
310
+ GATM (HGNC:4175),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
311
+ GATM (HGNC:4175),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
312
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
313
+ GATM (HGNC:4175),BP4,Supporting,"REVEL score <0.15 for missense variants.
314
+
315
+
316
+ In frame deletion or insertion predicted benign by PROVEAN and MutationTaster.
317
+
318
+
319
+ No predicted impact on splicing by SpliceAI and varSEAK.
320
+ GATM specifications:
321
+
322
+
323
+ For missense changes, REVEL score <0.15.
324
+
325
+
326
+ For in frame insertions and deletions, use PROVEAN and Mutation Taster. Results must be consistent to count.
327
+
328
+
329
+ For non-canonical splice site variants, use SpliceAI (
330
+ https://spliceailookup.broadinstitute.org/
331
+ ) and varSEAK (
332
+ https://varseak.bio/
333
+ ) to assess the impact of variants that are not +/-1 or 2 canonical splice site variants. For SpliceAI, this criterion can be applied for scores <0.2, and for varSEAK class 1 and 2. If there is any evidence for possible creation of a cryptic splice site, this criterion should not be applied.",None
334
+ GATM (HGNC:4175),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
335
+ GATM (HGNC:4175),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
336
+ GATM (HGNC:4175),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
337
+ GATM (HGNC:4175),BP7,Supporting,A synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,None
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGATMVersion1_version=1.0.0.csv ADDED
@@ -0,0 +1,568 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ GATM (HGNC:4175),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ GATM (HGNC:4175),PVS1,Very Strong,"Nonsense-mediated decay predicted.
9
+
10
+
11
+
12
+
13
+ CCDS VCEP notes:
14
+ Loss of function (LOF) of GATM is a known mechanism of disease for guanidinoacetate methyltransferase deficiency (AGAT-D). There are examples of various LOF variants, including nonsense and frameshift, in GATM in individuals with AGAT-D (
15
+ https://databases.lovd.nl/shared/variants/GATM?search_var_status=%3D%22Marked%22%7C%3D%22Public%22
16
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
17
+ GATM specifications:
18
+
19
+
20
+ Nonsense and frameshift variants
21
+
22
+
23
+
24
+
25
+ All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 9) or the last 50 bases of the penultimate exon (exon 8, 3’ of c.1109). In that case, PVS1_Moderate will be applied.
26
+
27
+
28
+
29
+
30
+ Splice site variants (+1, +2, -1, -2)
31
+
32
+
33
+
34
+
35
+ All canonical splice site pairs in GATM are GT-AG.
36
+
37
+
38
+ For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
39
+
40
+
41
+ For the predicted in frame/out of frame consequences of exon skipping and assigned strength of PVS1, see Appendix 1.
42
+
43
+
44
+ If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
45
+
46
+
47
+ To apply PVS1, splice site variants must have no detectable nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. Otherwise, the PVS1 strength should be reduced accordingly.
48
+
49
+
50
+ Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
51
+
52
+
53
+
54
+
55
+ Initiator codon variants
56
+
57
+
58
+
59
+
60
+ To our knowledge, initiator codon variants have not been reported in GATM (01/2019) but may occur.
61
+
62
+
63
+ All initiator codon variants will meet PVS1_Moderate. The next in-frame methionine is at amino acid position 130 (based on NP_001473).
64
+
65
+
66
+
67
+
68
+ Deletions (single or multi exon)
69
+
70
+
71
+
72
+
73
+ If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
74
+
75
+
76
+ If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
77
+
78
+
79
+ If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
80
+
81
+
82
+ Appendix 1 can be used to predict the consequences of single exon deletions.
83
+
84
+
85
+
86
+
87
+ Duplications
88
+
89
+
90
+
91
+
92
+ Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",None
93
+ GATM (HGNC:4175),PVS1,Strong,"In frame loss of >10% of the protein.
94
+ CCDS VCEP notes:
95
+ Loss of function (LOF) of GATM is a known mechanism of disease for guanidinoacetate methyltransferase deficiency (AGAT-D). There are examples of various LOF variants, including nonsense and frameshift, in GATM in individuals with AGAT-D (
96
+ https://databases.lovd.nl/shared/variants/GATM?search_var_status=%3D%22Marked%22%7C%3D%22Public%22
97
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
98
+ GATM specifications:
99
+
100
+
101
+
102
+
103
+ Nonsense and frameshift variants
104
+
105
+
106
+
107
+
108
+ All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 9) or the last 50 bases of the penultimate exon (exon 8, 3’ of c.1109). In that case, PVS1_Moderate will be applied.
109
+
110
+
111
+
112
+
113
+ Splice site variants (+1, +2, -1, -2)
114
+
115
+
116
+
117
+
118
+ All canonical splice site pairs in GATM are GT-AG.
119
+
120
+
121
+ For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
122
+
123
+
124
+ For the predicted in frame/out of frame consequences of exon skipping and assigned strength of PVS1, see Appendix 1.
125
+
126
+
127
+ If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
128
+
129
+
130
+ To apply PVS1, splice site variants must have no detectable nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. Otherwise, the PVS1 strength should be reduced accordingly.
131
+
132
+
133
+ Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
134
+
135
+
136
+
137
+
138
+ Initiator codon variants
139
+
140
+
141
+
142
+
143
+ To our knowledge, initiator codon variants have not been reported in GATM (01/2019) but may occur.
144
+
145
+
146
+ All initiator codon variants will meet PVS1_Moderate. The next in-frame methionine is at amino acid position 130 (based on NP_001473).
147
+
148
+
149
+
150
+
151
+ Deletions (single or multi exon)
152
+
153
+
154
+
155
+
156
+ If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
157
+
158
+
159
+ If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
160
+
161
+
162
+ If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
163
+
164
+
165
+ Appendix 1 can be used to predict the consequences of single exon deletions.
166
+
167
+
168
+
169
+
170
+ Duplications
171
+
172
+
173
+
174
+
175
+ Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",Strength
176
+ GATM (HGNC:4175),PVS1,Moderate,"Initiator codon variant.
177
+
178
+
179
+ Single exon or larger deletion resulting in loss of <10% of the protein.
180
+
181
+
182
+
183
+
184
+ CCDS VCEP notes:
185
+ Loss of function (LOF) of GATM is a known mechanism of disease for guanidinoacetate methyltransferase deficiency (AGAT-D). There are examples of various LOF variants, including nonsense and frameshift, in GATM in individuals with AGAT-D (
186
+ https://databases.lovd.nl/shared/variants/GATM?search_var_status=%3D%22Marked%22%7C%3D%22Public%22
187
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
188
+ GATM specifications:
189
+
190
+
191
+ Nonsense and frameshift variants
192
+
193
+
194
+
195
+
196
+ All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 9) or the last 50 bases of the penultimate exon (exon 8, 3’ of c.1109). In that case, PVS1_Moderate will be applied.
197
+
198
+
199
+
200
+
201
+ Splice site variants (+1, +2, -1, -2)
202
+
203
+
204
+
205
+
206
+ All canonical splice site pairs in GATM are GT-AG.
207
+
208
+
209
+ For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants.
210
+
211
+
212
+ For the predicted in frame/out of frame consequences of exon skipping and assigned strength of PVS1, see Appendix 1.
213
+
214
+
215
+ If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
216
+
217
+
218
+ To apply PVS1, splice site variants must have no detectable nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. Otherwise, the PVS1 strength should be reduced accordingly.
219
+
220
+
221
+ Non-canonical splice variants, such as +3 or -3, will not meet PVS1, but could meet PS3 and/or PP3 criteria.
222
+
223
+
224
+
225
+
226
+ Initiator codon variants
227
+
228
+
229
+
230
+
231
+ To our knowledge, initiator codon variants have not been reported in GATM (01/2019) but may occur.
232
+
233
+
234
+ All initiator codon variants will meet PVS1_Moderate. The next in-frame methionine is at amino acid position 130 (based on NP_001473).
235
+
236
+
237
+
238
+
239
+ Deletions (single or multi exon)
240
+
241
+
242
+
243
+
244
+ If a single or multi-exon deletion results in an out of frame consequence, use PVS1 unless not predicted to undergo NMD. If not predicted to undergo NMD, use PVS1_Strong if >10% of the protein is predicted to be removed, and use PVS1_Moderate if <10% of the protein is predicted to be removed.
245
+
246
+
247
+ If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. Use PVS1_Strong if more than 10% of the protein is removed and PVS1_Moderate if <10% of the protein is removed.
248
+
249
+
250
+ If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
251
+
252
+
253
+ Appendix 1 can be used to predict the consequences of single exon deletions.
254
+
255
+
256
+
257
+
258
+ Duplications
259
+
260
+
261
+
262
+
263
+ Use the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042) to assess the impact of duplications.",Strength
264
+ GATM (HGNC:4175),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
265
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
266
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
267
+ GATM (HGNC:4175),PS1,Strong,"CCDS VCEP notes:
268
+
269
+
270
+
271
+
272
+ This criterion is applicable as described.
273
+
274
+
275
+ If the variant is in the last 3 nucleotides of an exon, further analysis using splicing site prediction algorithms (see PP3) and data from the literature (if available) is required to investigate the impact on splicing.",None
276
+ GATM (HGNC:4175),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
277
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
278
+ GATM (HGNC:4175),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
279
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
280
+ GATM (HGNC:4175),PS3,Strong,"RT-PCR evidence of mis-splicing for non-canonical intronic variants with no evidence of normal splice products.
281
+
282
+
283
+
284
+
285
+ GATM specifications:
286
+ Any variant meeting the description for either in vitro expression or splicing assays can meet PS3 at the strengths given below. If a variant meets the description for both e.g., a splice site variant with no evidence of abnormal splicing and deficient AGAT activity in vitro, PS3 must only be counted once.
287
+ In vitro expression
288
+ AGAT activity data from an in vitro assay in which GATM variants were overexpressed in HeLa cells has been published (DesRoches et al, 2016; PMID 27233232). Any variant with AGAT activity at or below 15% of normal in this paper meets PS3_Supporting (see Appendix 2 for further details on AGAT functional assays).
289
+ Splicing assays
290
+ For non-canonical splicing variants, use PS3 if there is RT-PCR and/or RNA sequencing evidence demonstrating only abnormal splice products, with no evidence of normal splicing.
291
+
292
+
293
+
294
+
295
+ Evidence of abnormal splicing includes transcripts of alternative length or with specific intron or exon inclusion/exclusion. These studies can be performed on mRNA extracted from patient-derived cells, or by inserting the mutant genomic DNA into plasmid vectors and introducing these into human or other mammalian host cells.
296
+
297
+
298
+ Note that in patients who are compound heterozygotes for a splicing variant and another variants type that does not disrupt splicing, such as a missense variant, evidence of normal splicing is expected. However, the presence of normal splice products could complicate the assessment of the impact of the splice variant. Therefore, if there is any evidence of normal splice products, either when using RNA from patient cells or in an in vitro expression system, use PS3_Supporting.
299
+
300
+
301
+ PP3 may also be used for non-canonical splice variants meeting PS3 or PS3_Supporting.",Disease-specific
302
+ GATM (HGNC:4175),PS3,Supporting,"<15% control activity when variant is expressed in HeLa cells, as reported in PMID 27233232.
303
+
304
+
305
+ RT-PCR evidence of mis-splicing for non-canonical intronic variants with evidence of normal splice products.
306
+
307
+
308
+
309
+
310
+ GATM specifications:
311
+ Any variant meeting the description for either in vitro expression or splicing assays can meet PS3 at the strengths given below. If a variant meets the description for both e.g., a splice site variant with no evidence of abnormal splicing and deficient AGAT activity in vitro, PS3 must only be counted once.
312
+ In vitro expression
313
+ AGAT activity data from an in vitro assay in which GATM variants were overexpressed in HeLa cells has been published (DesRoches et al, 2016; PMID 27233232). Any variant with AGAT activity at or below 15% of normal in this paper meets PS3_Supporting (see Appendix 2 for further details on AGAT functional assays).
314
+ Splicing assays
315
+ For non-canonical splicing variants, use PS3 if there is RT-PCR and/or RNA sequencing evidence demonstrating only abnormal splice products, with no evidence of normal splicing.
316
+
317
+
318
+
319
+
320
+ Evidence of abnormal splicing includes transcripts of alternative length or with specific intron or exon inclusion/exclusion. These studies can be performed on mRNA extracted from patient-derived cells, or by inserting the mutant genomic DNA into plasmid vectors and introducing these into human or other mammalian host cells.
321
+
322
+
323
+ Note that in patients who are compound heterozygotes for a splicing variant and another variants type that does not disrupt splicing, such as a missense variant, evidence of normal splicing is expected. However, the presence of normal splice products could complicate the assessment of the impact of the splice variant. Therefore, if there is any evidence of normal splice products, either when using RNA from patient cells or in an in vitro expression system, use PS3_Supporting.
324
+
325
+
326
+ PP3 may also be used for non-canonical splice variants meeting PS3 or PS3_Supporting.",Disease-specific
327
+ GATM (HGNC:4175),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
328
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
329
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",NA
330
+ GATM (HGNC:4175),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
331
+ GATM (HGNC:4175),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
332
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
333
+ GATM (HGNC:4175),PM2,Supporting,"Allele frequency <0.000055 (<0.0055%) in all populations in gnomAD.
334
+ CCDS VCEP notes:
335
+ It is acceptable for a GATM variant to be present in controls, if heterozygous, because AGAT-D is a recessive disorder. Homozygotes should not be seen in a population database, such as gnomAD, because the penetrance of this condition in individuals with biallelic pathogenic variants is expected to be 100%.
336
+ GATM specifications:
337
+
338
+
339
+ All subpopulations in gnomAD must have a maximum allele frequency less than 0.000055 (based on the prevalence of the most common suspected pathogenic variants, c.484+1G>T and p.Arg169Ter) (see Appendix 3).
340
+ Note – PM2 will NOT be used at moderate strength; PM2 will only be applied as a Supporting criterion.
341
+
342
+
343
+ If homozygotes are observed, the variant will meet BS2 (assuming 100% penetrance for an individual with 2 pathogenic variants in trans).","Strength,Disease-specific"
344
+ GATM (HGNC:4175),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
345
+ Note: This requires testing of parents (or offspring) to determine phase.",
346
+ GATM (HGNC:4175),PM3,Very Strong,"Consult specifications for assigning strength of evidence for PM3.
347
+ GATM specifications:
348
+
349
+
350
+
351
+
352
+ Following SVI guidance for PM3 (
353
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
354
+ ), use the scoring system below.
355
+
356
+
357
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
358
+ GATM (HGNC:4175),PM3,Strong,"Consult specifications for assigning strength of evidence for PM3.
359
+ GATM specifications:
360
+
361
+
362
+
363
+
364
+ Following SVI guidance for PM3 (
365
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
366
+ ), use the scoring system below.
367
+
368
+
369
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
370
+ GATM (HGNC:4175),PM3,Moderate,"Consult specifications for assigning strength of evidence for PM3.
371
+ GATM specifications:
372
+
373
+
374
+
375
+
376
+ Following SVI guidance for PM3 (
377
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
378
+ ), use the scoring system below.
379
+
380
+
381
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",None
382
+ GATM (HGNC:4175),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
383
+ GATM (HGNC:4175),PM4,Moderate,"In frame deletion/insertions of two or more amino acids but less than one exon
384
+ CCDS VCEP notes:
385
+ • Stop loss variants in GATM have not been reported, as far as we are aware.
386
+ GATM specifications:
387
+ Use this rule “as is” for in frame deletions and insertions of 2 or more amino acids, but downgrade to PM4_Supporting for single amino acid deletions and insertions.",None
388
+ GATM (HGNC:4175),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
389
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
390
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
391
+ GATM (HGNC:4175),PM5,Moderate,"GATM specifications:
392
+
393
+
394
+
395
+
396
+ If the pathogenicity of another missense change at the same amino acid residue is unknown, determine its pathogenicity using these specifications in order to determine if this criterion can be used. If the variant is pathogenic, use PM5. If the variant is likely pathogenic, use PM5_Supporting.",None
397
+ GATM (HGNC:4175),PM5,Supporting,Missense change at an amino acid residue where a different missense change determined to be likely pathogenic has been seen before.,Strength
398
+ GATM (HGNC:4175),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
399
+ GATM (HGNC:4175),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
400
+ Note: May be used as stronger evidence with increasing segregation data.",NA
401
+ GATM (HGNC:4175),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
402
+ GATM (HGNC:4175),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
403
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
404
+ GATM (HGNC:4175),PP3,Moderate,"Non-canonical splice site variant predicted to be more deleterious than a previously observed pathogenic variant at the same nucleotide.
405
+ GATM specifications:
406
+
407
+
408
+ For missense changes, those with a REVEL score more than 0.75 will meet PP3.
409
+
410
+
411
+ For in frame insertions and deletions, use PROVEAN and Mutation Taster. Results must be consistent to count.
412
+
413
+
414
+ For non-canonical splice site variants (e.g., +3, -3), use SpliceAI ((
415
+ https://spliceailookup.broadinstitute.org/
416
+ ) and varSEAK (
417
+ https://varseak.bio/
418
+ ). Results must be consistent to apply this criterion.
419
+
420
+
421
+ For SpliceAI, any donor loss or acceptor loss with a score >0.5. For varSEAK, any variant with splicing class 4 or 5. Evidence for creation of a cryptic splice site should also be assessed.
422
+
423
+
424
+ Do not apply this rule for canonical splice site changes meeting PVS1.
425
+
426
+
427
+ Upgrade to PP3_Moderate if a different non-canonical splice site variant at the same position is known to be pathogenic AND the newly observed variant is at least as deleterious as the previously observed variant, based on in silico prediction (Human Splicing Finder, MaxEntScan) e.g. variant being interpreted is +3C>G, previously reported variant is +3C>T and +3C>G is predicted to be more deleterious than +3C>T.",None
428
+ GATM (HGNC:4175),PP3,Supporting,"REVEL score >0.75 for missense variants.
429
+
430
+
431
+ In frame deletion or insertion predicted deleterious by PROVEAN and MutationTaster.
432
+
433
+
434
+ Predicted impact on splicing by SpliceAI and varSEAK.
435
+ GATM specifications:
436
+
437
+
438
+ For missense changes, those with a REVEL score more than 0.75 will meet PP3.
439
+
440
+
441
+ For in frame insertions and deletions, use PROVEAN and Mutation Taster. Results must be consistent to count.
442
+
443
+
444
+ For non-canonical splice site variants (e.g., +3, -3), use SpliceAI ((
445
+ https://spliceailookup.broadinstitute.org/
446
+ ) and varSEAK (
447
+ https://varseak.bio/
448
+ ). Results must be consistent to apply this criterion.
449
+
450
+
451
+ For SpliceAI, any donor loss or acceptor loss with a score >0.5. For varSEAK, any variant with splicing class 4 or 5. Evidence for creation of a cryptic splice site should also be assessed.
452
+
453
+
454
+ Do not apply this rule for canonical splice site changes meeting PVS1.
455
+
456
+
457
+ Upgrade to PP3_Moderate if a different non-canonical splice site variant at the same position is known to be pathogenic AND the newly observed variant is at least as deleterious as the previously observed variant, based on in silico prediction (Human Splicing Finder, MaxEntScan) e.g. variant being interpreted is +3C>G, previously reported variant is +3C>T and +3C>G is predicted to be more deleterious than +3C>T.",None
458
+ GATM (HGNC:4175),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
459
+ GATM (HGNC:4175),PP4,Strong,"Any combination of the data listed for PP4_Moderate and PP4.
460
+ (see points scheme in main document)
461
+ GATM specifications:
462
+ The following table shows the weight used for PP4 depending upon the data that is available. Assign points for each type of data present, and then add up the points.
463
+ PP4 = 1-2 points
464
+ PP4_Moderate = 3 points
465
+ PP4_Strong = 4 points
466
+ Two or more data types are recommended to reach moderate and required to reach strong.
467
+
468
+
469
+
470
+
471
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
472
+
473
+
474
+ For PP4 to be applied at strong, full GATM gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
475
+ GATM (HGNC:4175),PP4,Moderate,"Significantly decreased creatine peak on 1H-MRS.
476
+
477
+
478
+ GATM activity <5% normal in fibroblasts.
479
+ GATM specifications:
480
+ The following table shows the weight used for PP4 depending upon the data that is available. Assign points for each type of data present, and then add up the points.
481
+ PP4 = 1-2 points
482
+ PP4_Moderate = 3 points
483
+ PP4_Strong = 4 points
484
+ Two or more data types are recommended to reach moderate and required to reach strong.
485
+
486
+
487
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
488
+
489
+
490
+ For PP4 to be applied at strong, full GATM gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Strength
491
+ GATM (HGNC:4175),PP4,Supporting,"Low GAA in plasma and low to low-normal creatine in plasma.
492
+ GATM specifications:
493
+ The following table shows the weight used for PP4 depending upon the data that is available. Assign points for each type of data present, and then add up the points.
494
+ PP4 = 1-2 points
495
+ PP4_Moderate = 3 points
496
+ PP4_Strong = 4 points
497
+ Two or more data types are recommended to reach moderate and required to reach strong.
498
+
499
+
500
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
501
+
502
+
503
+ For PP4 to be applied at strong, full GATM gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
504
+ GATM (HGNC:4175),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
505
+ GATM (HGNC:4175),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
506
+ GATM (HGNC:4175),BA1,Stand Alone,"Allele frequency >0.0005 (0.05%) in gnomAD in any continental population in gnomAD with >2000 alleles.
507
+ GATM specifications:
508
+
509
+
510
+
511
+
512
+ Any variant with a frequency >0.0005 (max allelic contribution = 100% and max genetic contribution = 100% based on estimated prevalence of 1 in 3,450,000 (PMID 27233232), and penetrance of 100%) in a continental population with >2000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383)(See Appendix 3).
513
+
514
+
515
+ Use the highest population minor allele frequency (MAF) in any given continental population with >2,000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383).",Disease-specific
516
+ GATM (HGNC:4175),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
517
+ GATM (HGNC:4175),BS1,Strong,"Allele frequency >0.0001 (0.01%) in gnomAD in any continental population in gnomAD with >2000 alleles.
518
+ GATM specifications:
519
+
520
+
521
+
522
+
523
+ Any variant with a frequency >0.0001 (max allelic contribution = 25% and max genetic contribution = 100% based on estimated prevalence of 1 in 3,450,000, (PMID 27233232), and penetrance of 100%) in a continental population with >2000 (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383) (see Appendix 3).
524
+
525
+
526
+ Use the highest population minor allele frequency (MAF) in any given continental population with >2,000 alleles (European non-Finnish, African, East Asian, South Asian, Latino) (PMID 30311383).",Disease-specific
527
+ GATM (HGNC:4175),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
528
+ GATM (HGNC:4175),BS2,Strong,Observed in the homozygous state in a healthy adult.,Disease-specific
529
+ GATM (HGNC:4175),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
530
+ GATM (HGNC:4175),BS3,Supporting,"30% normal GAA activity when the variant is expressed in a heterologous cell type.
531
+ GATM specifications:
532
+ In vitro assays
533
+ In vitro assays in which a variant is expressed in AGAT-deficient cultured cells (e.g. AGAT-deficient fibroblasts) or in-fusion High-Fidelity cloning of GATM transcript and site directed mutagenesis to generate missense variant overexpressed in HeLa cells and measurement of AGAT activity in cells for wild-type and missense variant. Any variant with enzyme activity at or above 30% of normal in DesRoches et al, 2016, PMID 27233232, meets BS3_Supporting.","Strength,Disease-specific"
534
+ GATM (HGNC:4175),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
535
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
536
+ GATM (HGNC:4175),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
537
+ GATM (HGNC:4175),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
538
+ GATM (HGNC:4175),BP2,Supporting,"Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder; or observed in cis with a pathogenic variant in any inheritance pattern.
539
+ GATM specifications:
540
+ Observed in cis with a pathogenic variant (to take AR inheritance into account).",None
541
+ GATM (HGNC:4175),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
542
+ GATM (HGNC:4175),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
543
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
544
+ GATM (HGNC:4175),BP4,Supporting,"REVEL score <0.15 for missense variants.
545
+
546
+
547
+ In frame deletion or insertion predicted benign by PROVEAN and MutationTaster.
548
+
549
+
550
+ No predicted impact on splicing by SpliceAI and varSEAK.
551
+ GATM specifications:
552
+
553
+
554
+ For missense changes, REVEL score <0.15.
555
+
556
+
557
+ For in frame insertions and deletions, use PROVEAN and Mutation Taster. Results must be consistent to count.
558
+
559
+
560
+ For non-canonical splice site variants, use SpliceAI (
561
+ https://spliceailookup.broadinstitute.org/
562
+ ) and varSEAK (
563
+ https://varseak.bio/
564
+ ) to assess the impact of variants that are not +/-1 or 2 canonical splice site variants. For SpliceAI, this criterion can be applied for scores <0.2, and for varSEAK class 1 and 2. If there is any evidence for possible creation of a cryptic splice site, this criterion should not be applied.",None
565
+ GATM (HGNC:4175),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
566
+ GATM (HGNC:4175),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
567
+ GATM (HGNC:4175),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
568
+ GATM (HGNC:4175),BP7,Supporting,A synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,None
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforGATMVersion2.0.0_version=2.0.0.csv ADDED
@@ -0,0 +1,272 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ GATM (HGNC:4175),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ GATM (HGNC:4175),PVS1,Very Strong,"Nonsense-mediated decay predicted.
9
+
10
+
11
+ CCDS VCEP notes:
12
+ Loss of function (LOF) of GATM is a known mechanism of disease for arginine:glycine amidinotransferase deficiency (AGAT-D). There are examples of various LOF variants, including nonsense and frameshift, in GATM in individuals with AGAT-D (
13
+ https://databases.lovd.nl/shared/variants/GATM/unique
14
+ ). The specifications below are based on the PVS1 decision tree (see Appendix 1) (based on Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
15
+
16
+
17
+ GATM specifications:
18
+ Nonsense and frameshift variants
19
+ * All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 9) or the last 50 bases of the penultimate exon (exon 8, 3’ of c.1109). In that case, please refer to the PVS1 flowchart for guidance on PVS1 weight (Appendix 1).
20
+
21
+
22
+ Splice site variants (+1, +2, -1, -2)
23
+ * All canonical splice site pairs in GATM are GT-AG.
24
+ * For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants. However, to apply PVS1 at the very strong level, splice site variants must have no detectable nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing, as predicted by SpliceAI. Otherwise, the PVS1 strength should be reduced accordingly.
25
+ * For the predicted in frame/out of frame consequences of exon skipping and considerations for strength of PVS1, see Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss).
26
+ * If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
27
+ * Non-canonical splice variants, such as +3 or -3, may also meet PVS1 - refer to Walker et al, PMID: 37352859.
28
+
29
+
30
+ Deletions (single or multi exon)
31
+ * If a single or multi-exon deletion results in an out of frame consequence, use PVS1 at the very strong level unless not predicted to undergo NMD. If not predicted to undergo NMD, please refer to Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss).
32
+ * If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. For weight of PVS1, see Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss) .
33
+ * If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
34
+
35
+
36
+ Duplications
37
+ * To assess the impact of duplications, see Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss).",None
38
+ GATM (HGNC:4175),PVS1,Strong,"In frame loss of >10% of the protein.
39
+ CCDS VCEP notes:
40
+ Loss of function (LOF) of GATM is a known mechanism of disease for arginine:glycine amidinotransferase deficiency (AGAT-D). There are examples of various LOF variants, including nonsense and frameshift, in GATM in individuals with AGAT-D (
41
+ https://databases.lovd.nl/shared/variants/GATM/unique
42
+ ). The specifications below are based on the PVS1 decision tree (see Appendix 1) (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).
43
+
44
+
45
+ GATM specifications:
46
+ Nonsense and frameshift variants
47
+ * All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 9) or the last 50 bases of the penultimate exon (exon 8, 3’ of c.1109). In that case, please refer to the PVS1 flowchart for guidance on PVS1 weight (Appendix 1).
48
+
49
+
50
+ Splice site variants (+1, +2, -1, -2)
51
+ * All canonical splice site pairs in GATM are GT-AG.
52
+ * For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants. However, it is recommended to use SpliceAI to look for nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. 
53
+ * For considerations for strength at which PVS1 may be applied see Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss) .
54
+ * If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
55
+ * Non-canonical splice variants, such as +3 or -3, may also meet PVS1 - refer to Walker et al, PMID: 37352859.
56
+
57
+
58
+ Deletions (single or multi exon)
59
+ * If a single or multi-exon deletion results in an out of frame consequence, use PVS1 at the very strong level unless not predicted to undergo NMD. If not predicted to undergo NMD, please refer to Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss).
60
+ * If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. For weight of PVS1, see Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss) .
61
+ * If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
62
+
63
+
64
+ Duplications
65
+ * To assess the impact of duplications, see Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss).",Strength
66
+ GATM (HGNC:4175),PVS1,Moderate,"Single exon or larger deletion resulting in loss of <10% of the protein, and initiator codon variants.
67
+
68
+
69
+ CCDS VCEP notes:
70
+ Loss of function (LOF) of GATM is a known mechanism of disease for arginine:glycine amidinotransferase deficiency (AGAT-D). There are examples of various LOF variants, including nonsense and frameshift, in GATM in individuals with AGAT-D (
71
+ https://databases.lovd.nl/shared/variants/GATM/unique
72
+ ). The specifications below are based on the PVS1 decision tree (Figure 1, Abou Tayoun et al, 2018, PMID 30192042).  
73
+
74
+
75
+ GATM specifications:
76
+
77
+
78
+ Nonsense and frameshift variants
79
+ * All nonsense and frameshift variants will meet PVS1 unless a premature termination codon is predicted to be missed by nonsense-mediated decay (NMD) because it is located in the last exon (exon 9) or the last 50 bases of the penultimate exon (exon 8, 3’ of c.1109). In that case, please refer to the PVS1 flowchart for guidance on PVS1 weight (Appendix 1).
80
+
81
+
82
+ Splice site variants (+1, +2, -1, -2)
83
+ * All canonical splice site pairs in GATM are GT-AG.
84
+ * For any canonical splice site variant (+1, +2, -1, -2), the exon immediately adjacent to the variant is predicted to be skipped i.e. upstream exon skipped for canonical donor splice site variants and downstream exon skipped for canonical acceptor splice site variants. However, it is recommended to look for nearby (+/- 20 nucleotides) strong consensus splice sequence that may reconstitute in-frame splicing. 
85
+ * For considerations for strength at which PVS1 may be applied see Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss).
86
+ * If this criterion is applied, PP3 (in silico splice site prediction tools) should not be used.
87
+ * Non-canonical splice variants, such as +3 or -3, may also meet PVS1 (refer to Walker et al, PMID: 37352859).
88
+
89
+
90
+ Initiator codon variants
91
+ * All initiator codon variants will meet PVS1_Moderate. The next in-frame methionine is at amino acid position 130 (based on NP_001473). 
92
+
93
+
94
+ Deletions (single or multi exon)
95
+ * If a single or multi-exon deletion results in an out of frame consequence, use PVS1 at the very strong level unless not predicted to undergo NMD. If not predicted to undergo NMD, please refer to Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss).
96
+ * If the consequence is in frame, the deletion must encompass one or more exons for PVS1 to apply. For weight of PVS1, see Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss) .
97
+ * If the in frame deletion is smaller than one exon, PVS1 does not apply; consider using PM4.
98
+
99
+
100
+ Duplications
101
+ * To assess the impact of duplications, see Appendix 1 (PVS1 flowchart) and Appendix 2 (predicted impact of exon loss).",Strength
102
+ GATM (HGNC:4175),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
103
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
104
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
105
+ GATM (HGNC:4175),PS1,Strong,"This criterion is applicable as is for any variant resulting in the same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
106
+
107
+
108
+ If the variant is in the last 3 nucleotides of an exon, further analysis using splicing site prediction algorithms (see PP3) and data from the literature (if available) is required to investigate the impact on splicing.
109
+
110
+
111
+ PS1 may also be applied for splicing variants under specific circumstances (see Table 3 in PMID: 37352859).",General recommendation
112
+ GATM (HGNC:4175),PS1,Moderate,"This criterion is applicable as for any variant resulting in the same amino acid change as a previously established likely pathogenic variant regardless of nucleotide change.
113
+
114
+
115
+ If the variant is in the last 3 nucleotides of an exon, further analysis using splicing site prediction algorithms (see PP3) and data from the literature (if available) is required to investigate the impact on splicing.
116
+
117
+
118
+ PS1_Moderate may also be applied for splicing variants under specific circumstances (see Table 3 in PMID: 37352859).","General recommendation,Strength"
119
+ GATM (HGNC:4175),PS1,Supporting,PS1_Supporting may be applied for splicing variants under specific circumstances (see Table 3 in PMID: 37352859).,"General recommendation,Strength"
120
+ GATM (HGNC:4175),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
121
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
122
+ GATM (HGNC:4175),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
123
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
124
+ GATM (HGNC:4175),PS3,Supporting,"AGAT activity data from an in vitro assay in which GATM variants were overexpressed in HeLa cells has been published (DesRoches et al, 2016; PMID 27233232). Any variant with AGAT activity at or below 15% of normal in this paper meets PS3_Supporting (see Appendix 3 for further details on AGAT functional assays).","Disease-specific,Strength"
125
+ GATM (HGNC:4175),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
126
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
127
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",NA
128
+ GATM (HGNC:4175),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
129
+ GATM (HGNC:4175),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
130
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
131
+ GATM (HGNC:4175),PM2,Supporting,"Allele frequency <0.000055 (<0.0055%) in all populations in gnomAD.
132
+
133
+
134
+ CCDS VCEP notes: It is acceptable for a GATM variant to be present in controls, if heterozygous, because AGAT-D is a recessive disorder. Homozygotes should not be seen in a population database, such as gnomAD, because the penetrance of this condition in individuals with biallelic pathogenic variants is expected to be 100%. 
135
+
136
+
137
+ GATM specifications:
138
+
139
+
140
+
141
+
142
+ All subpopulations in gnomAD must have a maximum allele frequency less than 0.000055 (based on the prevalence of the most common suspected pathogenic variants, c.484+1G>T and p.Arg169Ter) (see Appendix 4). Use the current version recommended by SVI; version number will be stated in classification summary.
143
+
144
+
145
+ Note – PM2 will NOT be used at moderate strength; PM2 will only be applied as a Supporting criterion.
146
+
147
+
148
+ If homozygotes are observed, the variant will meet BS2 (assuming 100% penetrance for an individual with 2 pathogenic variants in trans).","Disease-specific,Strength"
149
+ GATM (HGNC:4175),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
150
+ Note: This requires testing of parents (or offspring) to determine phase.",
151
+ GATM (HGNC:4175),PM3,Very Strong,"Follow SVI guidance for PM3 (
152
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
153
+ ).
154
+
155
+
156
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
157
+ GATM (HGNC:4175),PM3,Strong,"Follow SVI guidance for PM3 (
158
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
159
+ ).
160
+
161
+
162
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
163
+ GATM (HGNC:4175),PM3,Moderate,"Follow SVI guidance for PM3 (
164
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
165
+ ).
166
+
167
+
168
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",None
169
+ GATM (HGNC:4175),PM3,Supporting,"Follow SVI guidance for PM3 (
170
+ https://clinicalgenome.org/site/assets/files/3717/svi_proposal_for_pm3_criterion_-_version_1.pdf
171
+ ).
172
+
173
+
174
+ Parental testing, or another appropriate molecular method (such as cloning each allele separately followed by sequencing), must have been performed in order to confirm that the variants are in trans if the patient is compound heterozygous.",Strength
175
+ GATM (HGNC:4175),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
176
+ GATM (HGNC:4175),PM4,Moderate,"CCDS VCEP notes: Stop loss variants in GATM have not been reported, as far as we are aware. 
177
+
178
+
179
+ GATM specifications: Use this rule “as is” for in frame deletions and insertions of 2 or more amino acids, but downgrade to PM4_Supporting for single amino acid deletions and insertions.",None
180
+ GATM (HGNC:4175),PM4,Supporting,Downgrade to PM4_Supporting for in frame deletion/insertion of a single amino acid.,Strength
181
+ GATM (HGNC:4175),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
182
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
183
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
184
+ GATM (HGNC:4175),PM5,Moderate,"If the pathogenicity of another missense change at the same amino acid residue is unknown, determine its pathogenicity using these specifications in order to determine if this criterion can be used. If the variant is pathogenic, use PM5. If the variant is likely pathogenic, use PM5_Supporting.",None
185
+ GATM (HGNC:4175),PM5,Supporting,Missense change at an amino acid residue where a different missense change determined to be likely pathogenic has been seen before.,Strength
186
+ GATM (HGNC:4175),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
187
+ GATM (HGNC:4175),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
188
+ Note: May be used as stronger evidence with increasing segregation data.",NA
189
+ GATM (HGNC:4175),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
190
+ GATM (HGNC:4175),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
191
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
192
+ GATM (HGNC:4175),PP3,Moderate,"Missense variant with a REVEL score >0.773 (based on guidance from Pejaver et al, 2022, PMID: 36413997).","General recommendation,Strength"
193
+ GATM (HGNC:4175),PP3,Supporting,"Missense variant with a REVEL score 0.644-0.773 (based on guidance from Pejaver et al, 2022, PMID: 36413997).
194
+
195
+
196
+ In frame deletion or insertion predicted deleterious by PROVEAN and MutationTaster. Results must be consistent to count.
197
+
198
+
199
+ For non-canonical splice site variants (e.g., +3, -3), predict the impact on splicing by using SpliceAI,
200
+ https://spliceailookup.broadinstitute.org/
201
+ , and apply PP3 for a score equal to or >0.2 (as indicated in PMID: 37352859, Table 1 and Figure 4). Assess the possibility of activation of cryptic splice sites.",General recommendation
202
+ GATM (HGNC:4175),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
203
+ GATM (HGNC:4175),PP4,Strong,"4 or more points based on any combination of the following. Two or more data types are required to meet Strong:
204
+ • Low urine guanidinoacetate with or without low or low normal creatine (1 point)
205
+ • Low plasma guanidinoacetate with or without low or low normal creatine (2 points)
206
+ • Significantly decreased creatine peak in brain magnetic resonance spectroscopy (3 points)
207
+ • AGAT enzyme activity <5% of normal (3 points)
208
+
209
+
210
+ * Variant must meet PM2_Supporting for PP4 to apply at any strength.
211
+ * For PP4 to be applied at strong, full GATM gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
212
+ GATM (HGNC:4175),PP4,Moderate,"3 points based on any combination of the following. Two or more data types are recommended to reach moderate:
213
+ • Low urine guanidinoacetate with or without low or low normal creatine (1 point)
214
+ • Low plasma guanidinoacetate with or without low or low normal creatine (2 points)
215
+ • Significantly decreased creatine peak in brain magnetic resonance spectroscopy (3 points)
216
+ • AGAT enzyme activity <5% of normal (3 points)
217
+
218
+
219
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.",Strength
220
+ GATM (HGNC:4175),PP4,Supporting,"1-2 points based on:
221
+ • Low urine guanidinoacetate with or without low or low normal creatine (1 point)
222
+ • Low plasma guanidinoacetate with or without low or low normal creatine (2 points)
223
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.",Disease-specific
224
+ GATM (HGNC:4175),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
225
+ GATM (HGNC:4175),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
226
+ GATM (HGNC:4175),BA1,Stand Alone,"GrpMax >0.0005 (0.05%) in gnomAD
227
+
228
+
229
+
230
+
231
+ Any variant with a GrpMax (lower bound 95%ile) >0.0005 in gnomAD. Use the current version recommended by SVI; version number will be stated in classification summary. 
232
+
233
+
234
+ Threshold based on (max allelic contribution = 100% and max genetic contribution = 100% based on estimated prevalence of 1 in 3,450,000 (PMID 27233232), and penetrance of 100%) (PMID 30311383) (see Appendix 3)",Disease-specific
235
+ GATM (HGNC:4175),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
236
+ GATM (HGNC:4175),BS1,Strong,"GrpMax >0.0001 (0.01%) in gnomAD
237
+
238
+
239
+
240
+
241
+ Any variant with a GrpMax (lower bound 95%ile) >0.0001 in gnomAD. Use the current version recommended by SVI; version number will be stated in classification summary.
242
+
243
+
244
+  Threshold based on max allelic contribution = 25% and max genetic contribution = 100% based on estimated prevalence of 1 in 3,450,000, (PMID 27233232), and penetrance of 100%) (PMID 30311383) (see Appendix 3).",Disease-specific
245
+ GATM (HGNC:4175),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
246
+ GATM (HGNC:4175),BS2,Strong,Observed in the homozygous state in a healthy adult.,Disease-specific
247
+ GATM (HGNC:4175),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
248
+ GATM (HGNC:4175),BS3,Supporting,"In vitro assays in which a variant is expressed in AGAT-deficient cultured cells (e.g. AGAT-deficient fibroblasts) or in-fusion High-Fidelity cloning of GATM transcript and site directed mutagenesis to generate missense variant overexpressed in HeLa cells and measurement of AGAT activity in cells for wild-type and missense variant. Any variant with enzyme activity at or above 30% of normal in DesRoches et al, 2016, PMID 27233232, meets BS3_Supporting.","Disease-specific,Strength"
249
+ GATM (HGNC:4175),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
250
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
251
+ GATM (HGNC:4175),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
252
+ GATM (HGNC:4175),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
253
+ GATM (HGNC:4175),BP2,Supporting,Observed in cis with a pathogenic variant (to take AR inheritance for AGAT deficiency into account).,None
254
+ GATM (HGNC:4175),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
255
+ GATM (HGNC:4175),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
256
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
257
+ GATM (HGNC:4175),BP4,Supporting,"REVEL score <0.29 for missense variants (based on guidance from Pejaver et al, 2022, PMID: 36413997).
258
+
259
+
260
+ In frame deletion or insertion predicted benign by MutPredIndel and MutationTaster.
261
+
262
+
263
+ No predicted impact on splicing by SpliceAI,
264
+ https://spliceailookup.broadinstitute.org/
265
+ , based on a score <0.1 (as indicated in PMID: 37352859, Table 1 and Figure 4).","General recommendation,None"
266
+ GATM (HGNC:4175),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
267
+ GATM (HGNC:4175),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
268
+ GATM (HGNC:4175),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
269
+ GATM (HGNC:4175),BP7,Strong,"Experimental evidence, such as RT-PCR, shows no impact on splicing. Follow the decision tree outlined in Figure 5, Walker et al, 2023, PMID: 37352859. Note that splicing may appear normal in compound heterozygous patients with one allele that is degraded by nonsense-mediated decay as the results of a frameshift and premature termination codon due to splicing defect. Therefore. caution must be used in asessing the data prior to applying this code.","General recommendation,Strength"
270
+ GATM (HGNC:4175),BP7,Supporting,"A synonymous (silent) variant OR an intronic variant at or beyond positions +7 and -21, for which SpliceAI,
271
+ https://spliceailookup.broadinstitute.org/
272
+ , predicts no impact on splicing (score <0.1) (see Walker et al, 2023, PMID: 37352859).",None
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSLC6A8Version1.1.0_version=1.1.0.csv ADDED
@@ -0,0 +1,209 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ SLC6A8 (HGNC:11055),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ SLC6A8 (HGNC:11055),PVS1,Very Strong,"Loss of function is a known mechanism of disease for Creatine Transporter Deficiency.
9
+
10
+
11
+ Specifications are based on the decision tree as outlined in Tayoun etal, 2018 (Hum Mutat. 2018 Nov;39(11):1517-1524; PMID: 30192042) SLC6A8: PVS1, at appropriate strength, is applicable as described in Abou Tayoun et al, 2018 (PMID: 30192042)",None
12
+ SLC6A8 (HGNC:11055),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
13
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
14
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
15
+ SLC6A8 (HGNC:11055),PS1,Strong,PS1 is applicable as described.,None
16
+ SLC6A8 (HGNC:11055),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
17
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
18
+ SLC6A8 (HGNC:11055),PS2,Strong,"Note: Confirmation of paternity in females only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.
19
+
20
+
21
+ X-linked disorder. Only maternity needs to be confirmed.","Disease-specific,None"
22
+ SLC6A8 (HGNC:11055),PS2,Moderate,Newly hemizygous male with the variant identified de novo in the mother with no family history of other affected males.,"Disease-specific,Strength"
23
+ SLC6A8 (HGNC:11055),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
24
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
25
+ SLC6A8 (HGNC:11055),PS3,Strong,"RT-PCR evidence of mis-splicing for non-canonical intronic variants.
26
+
27
+
28
+ For non-canonical splicing variants, RT-PCR evidence demonstrating transcripts of alternative length or specific intron or exon inclusion/exclusion. These studies can be performed in patient derived cells, by placing the mutant genomic DNA into plasmid vectors, or by over-expressing mutant transcript. Assays should demonstrate defective splicing with RT-PCR analysis or RNA sequencing to confirm alternative transcripts.",Disease-specific
29
+ SLC6A8 (HGNC:11055),PS3,Supporting,"Creatine transport activity <10% wild type with less than or equal to 125uM creatine used in SLC6A8 deficient fibroblasts
30
+
31
+
32
+ RT-PCR evidence of mis-splicing for non-canonical intronic variants with evidence of normal splice products.
33
+ For non-canonical splicing variants, RT-PCR evidence demonstrating transcripts of alternative length or specific intron or exon inclusion/exclusion. These studies can be performed in patient derived cells, by placing the mutant genomic DNA into plasmid vectors, or by over-expressing mutant transcript. Assays should demonstrate defective splicing with RT-PCR analysis or RNA sequencing to confirm alternative transcripts.",Disease-specific
34
+ SLC6A8 (HGNC:11055),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
35
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
36
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
37
+ SLC6A8 (HGNC:11055),PS4,Very Strong,"4 independent male probands with elevated urine creatine/creatinine ratio on one occasion at minimum, in addition to any proband used for PP4.
38
+
39
+
40
+ Variant must meet PM2_Supporting criterion for PS4 to apply.",Strength
41
+ SLC6A8 (HGNC:11055),PS4,Strong,"3 independent male probands with elevated urine creatine/creatinine ratio on one occasion at minimum, in addition to any proband used for PP4.
42
+
43
+
44
+ Variant must meet PM2_Supporting criterion for PS4 to apply.",Disease-specific
45
+ SLC6A8 (HGNC:11055),PS4,Moderate,"2 independent male probands with elevated urine creatine/creatinine ratio on one occasion at minimum, in addition to any proband used for PP4.
46
+
47
+
48
+ Variant must meet PM2_Supporting criterion for PS4 to apply.",Strength
49
+ SLC6A8 (HGNC:11055),PS4,Supporting,"One independent male proband in addition to any proband used for PP4.
50
+
51
+
52
+ Variant must meet PM2_Supporting criterion for PS4 to apply.",Strength
53
+ SLC6A8 (HGNC:11055),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
54
+ SLC6A8 (HGNC:11055),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
55
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
56
+ SLC6A8 (HGNC:11055),PM2,Supporting,Absent/rare from controls in an ethnically-matched cohort population sample. Threshold: <0.00002 (0.002%) AND 0 hemizygotes in gnomAD.,Disease-specific
57
+ SLC6A8 (HGNC:11055),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
58
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
59
+ SLC6A8 (HGNC:11055),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
60
+ SLC6A8 (HGNC:11055),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,None
61
+ SLC6A8 (HGNC:11055),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
62
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
63
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
64
+ SLC6A8 (HGNC:11055),PM5,Moderate,Missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.,None
65
+ SLC6A8 (HGNC:11055),PM5,Supporting,Missense change at an amino acid residue where a different missense change determined to be likely pathogenic has been seen before.,Strength
66
+ SLC6A8 (HGNC:11055),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
67
+ SLC6A8 (HGNC:11055),PM6,Moderate,Variant identified as de novo in an affected male with the mother negative for the variant but maternity not confirmed.,No change
68
+ SLC6A8 (HGNC:11055),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
69
+ Note: May be used as stronger evidence with increasing segregation data.",
70
+ SLC6A8 (HGNC:11055),PP1,Strong,"3 affected segregations + 0 unaffected segregations OR
71
+
72
+
73
+ 2 affected segregations + 3 unaffected segregations",Strength
74
+ SLC6A8 (HGNC:11055),PP1,Moderate,2 affected segregations + 0 unaffected segregations.,Strength
75
+ SLC6A8 (HGNC:11055),PP1,Supporting,1 affected family member + 3 unaffected segregations.,Disease-specific
76
+ SLC6A8 (HGNC:11055),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
77
+ SLC6A8 (HGNC:11055),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
78
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
79
+ SLC6A8 (HGNC:11055),PP3,Supporting,"REVEL score >0.75 for missense variants
80
+
81
+
82
+ In frame deletion or insertion predicted deleterious by PROVEAN, MutPred indel, and MutationTaster.
83
+
84
+
85
+ Predicted impact on splicing by SpliceAI and varSEAK.",None
86
+ SLC6A8 (HGNC:11055),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
87
+ SLC6A8 (HGNC:11055),PP4,Strong,"4 or more points based on combinations of the following. 
88
+
89
+
90
+
91
+
92
+ Elevated urine creatine/creatinine ratio on one occasion (1 point)
93
+
94
+
95
+ Elevated urine creatine/creatinine ratio on more than one occasion (2 points)
96
+
97
+
98
+ Significantly decreased creatine peak, with absent guanidinoacetate peak, if reported (3 points)
99
+
100
+
101
+ Deficient creatine uptake in cultured fibroblasts (<10% of normal with <125uM creatine) (3 points)
102
+
103
+
104
+
105
+
106
+ Additional specifications:
107
+
108
+
109
+
110
+
111
+ Two or more data types are required for PP4_Strong.
112
+
113
+
114
+ An individual used to assign PP4, at any weight, cannot be also included for PS4 count. If multiple unrelated probands with the variant have been identified, it is recommended that the case with the highest PP4 points is assigned the appropriate weight for PP4, and the other cases are used for PS4.
115
+
116
+
117
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
118
+
119
+
120
+ For PP4 to be applied at strong, full SLC6A8 gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
121
+ SLC6A8 (HGNC:11055),PP4,Moderate,"3 or more points based on: 
122
+
123
+
124
+
125
+
126
+ Elevated urine creatine/creatinine ratio on one occasion (1 point)
127
+
128
+
129
+ Elevated urine creatine/creatinine ratio on more than one occasion (2 points)
130
+
131
+
132
+ Significantly decreased creatine peak, with absent guanidinoacetate peak, if reported (3 points)
133
+
134
+
135
+ Deficient creatine uptake in cultured fibroblasts (<10% of normal with <125uM creatine) (3 points)
136
+
137
+
138
+
139
+
140
+ Additional specifications:
141
+
142
+
143
+
144
+
145
+ Two or more data types are recommended for PP4_Moderate.
146
+
147
+
148
+ An individual used to assign PP4, at any weight, cannot be also included for PS4 count. If multiple unrelated probands with the variant have been identified, it is recommended that the case with the highest PP4 points is assigned the appropriate weight for PP4, and the other cases are used for PS4.
149
+
150
+
151
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.",Strength
152
+ SLC6A8 (HGNC:11055),PP4,Supporting,"1-2 or more points based on: 
153
+
154
+
155
+
156
+
157
+ Elevated urine creatine/creatinine ratio on one occasion (1 point)
158
+
159
+
160
+ Elevated urine creatine/creatinine ratio on more than one occasion (2 points)
161
+
162
+
163
+
164
+
165
+ Additional specifications:
166
+
167
+
168
+
169
+
170
+ An individual used to assign PP4, at any weight, cannot be also included for PS4 count. If multiple unrelated probands with the variant have been identified, it is recommended that the case with the highest PP4 points is assigned the appropriate weight for PP4, and the other cases are used for PS4.
171
+
172
+
173
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.",Disease-specific
174
+ SLC6A8 (HGNC:11055),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
175
+ SLC6A8 (HGNC:11055),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
176
+ SLC6A8 (HGNC:11055),BA1,Stand Alone,Allele frequency >0.0020 (0.2%) OR ≥10 hemizygotes in gnomAD,Disease-specific
177
+ SLC6A8 (HGNC:11055),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
178
+ SLC6A8 (HGNC:11055),BS1,Strong,Allele frequency > 0.0002 (0.02%) OR ≥ 5 hemizygotes in gnomAD,Disease-specific
179
+ SLC6A8 (HGNC:11055),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
180
+ SLC6A8 (HGNC:11055),BS2,Strong,Observed in the homozygous state in a healthy adult,None
181
+ SLC6A8 (HGNC:11055),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
182
+ SLC6A8 (HGNC:11055),BS3,Supporting,"Creatine transport assay demonstrating ≥50% normal transport activity using physiological creatine concentrations (≤125μM creatine).
183
+
184
+
185
+ RT-PCR evidence demonstrating no observable effect of splicing.
186
+
187
+
188
+ Expression assay demonstrating wild type transcript levels","Disease-specific,Strength"
189
+ SLC6A8 (HGNC:11055),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
190
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
191
+ SLC6A8 (HGNC:11055),BS4,Strong,Lack of segregation in affected members of a family.,None
192
+ SLC6A8 (HGNC:11055),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
193
+ SLC6A8 (HGNC:11055),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,NA
194
+ SLC6A8 (HGNC:11055),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
195
+ SLC6A8 (HGNC:11055),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
196
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
197
+ SLC6A8 (HGNC:11055),BP4,Supporting,"REVEL score <0.2 for missense variants
198
+
199
+
200
+ In frame deletion or insertion predicted benign by PROVEAN, MutPred indel, and MutationTaster.
201
+
202
+
203
+ No predicted impact on splicing by SpliceAI and varSEAK.",None
204
+ SLC6A8 (HGNC:11055),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
205
+ SLC6A8 (HGNC:11055),BP5,Supporting,"Variant found in a case with an alternate molecular basis for disease.
206
+ BP5 applicable as described; the case must have specific features of creatine transporter deficiency, such as low creatine on brain magnetic resonance spectroscopy, or elevated urine creatine in order to apply this criterion. Presence of developmental delay and seizures is not sufficient.",None
207
+ SLC6A8 (HGNC:11055),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
208
+ SLC6A8 (HGNC:11055),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
209
+ SLC6A8 (HGNC:11055),BP7,Supporting,A synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,None
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSLC6A8Version1.2.0_version=1.2.0.csv ADDED
@@ -0,0 +1,208 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ SLC6A8 (HGNC:11055),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ SLC6A8 (HGNC:11055),PVS1,Very Strong,"Loss of function is a known mechanism of disease for Creatine Transporter Deficiency.
9
+
10
+
11
+ Specifications are based on the decision tree as outlined in Tayoun etal, 2018 (Hum Mutat. 2018 Nov;39(11):1517-1524; PMID: 30192042) SLC6A8: PVS1, at appropriate strength, is applicable as described in Abou Tayoun et al, 2018 (PMID: 30192042)",None
12
+ SLC6A8 (HGNC:11055),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
13
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
14
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
15
+ SLC6A8 (HGNC:11055),PS1,Strong,PS1 is applicable as described.,"General recommendation,None"
16
+ SLC6A8 (HGNC:11055),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
17
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
18
+ SLC6A8 (HGNC:11055),PS2,Strong,"Note: Confirmation of paternity in females only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.
19
+
20
+
21
+ X-linked disorder. Only maternity needs to be confirmed.","Disease-specific,None"
22
+ SLC6A8 (HGNC:11055),PS2,Moderate,Newly hemizygous male with the variant identified de novo in the mother with no family history of other affected males.,"Disease-specific,Strength"
23
+ SLC6A8 (HGNC:11055),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
24
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
25
+ SLC6A8 (HGNC:11055),PS3,Strong,"RT-PCR evidence of mis-splicing for non-canonical intronic variants.
26
+
27
+
28
+ For non-canonical splicing variants, RT-PCR evidence demonstrating transcripts of alternative length or specific intron or exon inclusion/exclusion. These studies can be performed in patient derived cells, by placing the mutant genomic DNA into plasmid vectors, or by over-expressing mutant transcript. Assays should demonstrate defective splicing with RT-PCR analysis or RNA sequencing to confirm alternative transcripts.",Disease-specific
29
+ SLC6A8 (HGNC:11055),PS3,Supporting,"Creatine transport activity <10% wild type with less than or equal to 125uM creatine used in SLC6A8 deficient fibroblasts
30
+
31
+
32
+ RT-PCR evidence of mis-splicing for non-canonical intronic variants with evidence of normal splice products. For non-canonical splicing variants, RT-PCR evidence demonstrating transcripts of alternative length or specific intron or exon inclusion/exclusion. These studies can be performed in patient derived cells, by placing the mutant genomic DNA into plasmid vectors, or by over-expressing mutant transcript. Assays should demonstrate defective splicing with RT-PCR analysis or RNA sequencing to confirm alternative transcripts.","Disease-specific,Strength"
33
+ SLC6A8 (HGNC:11055),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
34
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
35
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
36
+ SLC6A8 (HGNC:11055),PS4,Very Strong,"4 independent male probands with elevated urine creatine/creatinine ratio on one occasion at minimum, in addition to any proband used for PP4.
37
+
38
+
39
+ Variant must meet PM2_Supporting criterion for PS4 to apply.",Strength
40
+ SLC6A8 (HGNC:11055),PS4,Strong,"3 independent male probands with elevated urine creatine/creatinine ratio on one occasion at minimum, in addition to any proband used for PP4.
41
+
42
+
43
+ Variant must meet PM2_Supporting criterion for PS4 to apply.",Disease-specific
44
+ SLC6A8 (HGNC:11055),PS4,Moderate,"2 independent male probands with elevated urine creatine/creatinine ratio on one occasion at minimum, in addition to any proband used for PP4.
45
+
46
+
47
+ Variant must meet PM2_Supporting criterion for PS4 to apply.",Strength
48
+ SLC6A8 (HGNC:11055),PS4,Supporting,"One independent male proband in addition to any proband used for PP4.
49
+
50
+
51
+ Variant must meet PM2_Supporting criterion for PS4 to apply.",Strength
52
+ SLC6A8 (HGNC:11055),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
53
+ SLC6A8 (HGNC:11055),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
54
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
55
+ SLC6A8 (HGNC:11055),PM2,Supporting,Absent/rare from controls in an ethnically-matched cohort population sample. Threshold: <0.00002 (0.002%) AND 0 hemizygotes in gnomAD.,Disease-specific
56
+ SLC6A8 (HGNC:11055),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
57
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
58
+ SLC6A8 (HGNC:11055),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
59
+ SLC6A8 (HGNC:11055),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,None
60
+ SLC6A8 (HGNC:11055),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
61
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
62
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
63
+ SLC6A8 (HGNC:11055),PM5,Moderate,Missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.,None
64
+ SLC6A8 (HGNC:11055),PM5,Supporting,Missense change at an amino acid residue where a different missense change determined to be likely pathogenic has been seen before.,Strength
65
+ SLC6A8 (HGNC:11055),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
66
+ SLC6A8 (HGNC:11055),PM6,Moderate,Variant identified as de novo in an affected male with the mother negative for the variant but maternity not confirmed.,No change
67
+ SLC6A8 (HGNC:11055),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
68
+ Note: May be used as stronger evidence with increasing segregation data.",
69
+ SLC6A8 (HGNC:11055),PP1,Strong,"3 affected segregations + 0 unaffected segregations OR
70
+
71
+
72
+ 2 affected segregations + 3 unaffected segregations",Strength
73
+ SLC6A8 (HGNC:11055),PP1,Moderate,2 affected segregations + 0 unaffected segregations.,Strength
74
+ SLC6A8 (HGNC:11055),PP1,Supporting,1 affected family member + 3 unaffected segregations.,Disease-specific
75
+ SLC6A8 (HGNC:11055),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
76
+ SLC6A8 (HGNC:11055),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
77
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
78
+ SLC6A8 (HGNC:11055),PP3,Supporting,"REVEL score >0.75 for missense variants
79
+
80
+
81
+ In frame deletion or insertion predicted deleterious by PROVEAN, MutPred indel, and MutationTaster.
82
+
83
+
84
+ Predicted impact on splicing by SpliceAI and varSEAK.",General recommendation
85
+ SLC6A8 (HGNC:11055),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
86
+ SLC6A8 (HGNC:11055),PP4,Strong,"4 or more points based on combinations of the following. 
87
+
88
+
89
+
90
+
91
+ Elevated urine creatine/creatinine ratio on one occasion (1 point)
92
+
93
+
94
+ Elevated urine creatine/creatinine ratio on more than one occasion (2 points)
95
+
96
+
97
+ Significantly decreased creatine peak, with absent guanidinoacetate peak, if reported (3 points)
98
+
99
+
100
+ Deficient creatine uptake in cultured fibroblasts (<10% of normal with <125uM creatine) (3 points)
101
+
102
+
103
+
104
+
105
+ Additional specifications:
106
+
107
+
108
+
109
+
110
+ Two or more data types are required for PP4_Strong.
111
+
112
+
113
+ An individual used to assign PP4, at any weight, cannot be also included for PS4 count. If multiple unrelated probands with the variant have been identified, it is recommended that the case with the highest PP4 points is assigned the appropriate weight for PP4, and the other cases are used for PS4.
114
+
115
+
116
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
117
+
118
+
119
+ For PP4 to be applied at strong, full SLC6A8 gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
120
+ SLC6A8 (HGNC:11055),PP4,Moderate,"3 or more points based on: 
121
+
122
+
123
+
124
+
125
+ Elevated urine creatine/creatinine ratio on one occasion (1 point)
126
+
127
+
128
+ Elevated urine creatine/creatinine ratio on more than one occasion (2 points)
129
+
130
+
131
+ Significantly decreased creatine peak, with absent guanidinoacetate peak, if reported (3 points)
132
+
133
+
134
+ Deficient creatine uptake in cultured fibroblasts (<10% of normal with <125uM creatine) (3 points)
135
+
136
+
137
+
138
+
139
+ Additional specifications:
140
+
141
+
142
+
143
+
144
+ Two or more data types are recommended for PP4_Moderate.
145
+
146
+
147
+ An individual used to assign PP4, at any weight, cannot be also included for PS4 count. If multiple unrelated probands with the variant have been identified, it is recommended that the case with the highest PP4 points is assigned the appropriate weight for PP4, and the other cases are used for PS4.
148
+
149
+
150
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.",Strength
151
+ SLC6A8 (HGNC:11055),PP4,Supporting,"1-2 or more points based on: 
152
+
153
+
154
+
155
+
156
+ Elevated urine creatine/creatinine ratio on one occasion (1 point)
157
+
158
+
159
+ Elevated urine creatine/creatinine ratio on more than one occasion (2 points)
160
+
161
+
162
+
163
+
164
+ Additional specifications:
165
+
166
+
167
+
168
+
169
+ An individual used to assign PP4, at any weight, cannot be also included for PS4 count. If multiple unrelated probands with the variant have been identified, it is recommended that the case with the highest PP4 points is assigned the appropriate weight for PP4, and the other cases are used for PS4.
170
+
171
+
172
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.",Disease-specific
173
+ SLC6A8 (HGNC:11055),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
174
+ SLC6A8 (HGNC:11055),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
175
+ SLC6A8 (HGNC:11055),BA1,Stand Alone,Allele frequency >0.0020 (0.2%) OR ≥10 hemizygotes in gnomAD,Disease-specific
176
+ SLC6A8 (HGNC:11055),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
177
+ SLC6A8 (HGNC:11055),BS1,Strong,Allele frequency > 0.0002 (0.02%) OR ≥ 5 hemizygotes in gnomAD,Disease-specific
178
+ SLC6A8 (HGNC:11055),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
179
+ SLC6A8 (HGNC:11055),BS2,Strong,"The variant is observed in ≥2 hemizygotes in gnomAD, or one or more hemizygous male(s) or homozygous female(s) who have documented normal urine creatine/creatine ratio.",None
180
+ SLC6A8 (HGNC:11055),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
181
+ SLC6A8 (HGNC:11055),BS3,Supporting,"Creatine transport assay demonstrating ≥50% normal transport activity using physiological creatine concentrations (≤125μM creatine).
182
+
183
+
184
+ RT-PCR evidence demonstrating no observable effect of splicing.
185
+
186
+
187
+ Expression assay demonstrating wild type transcript levels","Disease-specific,Strength"
188
+ SLC6A8 (HGNC:11055),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
189
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
190
+ SLC6A8 (HGNC:11055),BS4,Strong,Lack of segregation in affected members of a family.,None
191
+ SLC6A8 (HGNC:11055),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
192
+ SLC6A8 (HGNC:11055),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,NA
193
+ SLC6A8 (HGNC:11055),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
194
+ SLC6A8 (HGNC:11055),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
195
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
196
+ SLC6A8 (HGNC:11055),BP4,Supporting,"REVEL score <0.2 for missense variants
197
+
198
+
199
+ In frame deletion or insertion predicted benign by PROVEAN, MutPred indel, and MutationTaster.
200
+
201
+
202
+ No predicted impact on splicing by SpliceAI and varSEAK.",General recommendation
203
+ SLC6A8 (HGNC:11055),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
204
+ SLC6A8 (HGNC:11055),BP5,Supporting,"Variant found in a case with an alternate molecular basis for disease.
205
+ BP5 applicable as described; the case must have specific features of creatine transporter deficiency, such as low creatine on brain magnetic resonance spectroscopy, or elevated urine creatine in order to apply this criterion. Presence of developmental delay and seizures is not sufficient.",None
206
+ SLC6A8 (HGNC:11055),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
207
+ SLC6A8 (HGNC:11055),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
208
+ SLC6A8 (HGNC:11055),BP7,Supporting,A synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,"General recommendation,None"
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCerebralCreatineDeficiencySyndromesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSLC6A8Version1_version=1.0.0.csv ADDED
@@ -0,0 +1,190 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ SLC6A8 (HGNC:11055),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ SLC6A8 (HGNC:11055),PVS1,Very Strong,"Loss of function is a known mechanism of disease for Creatine Transporter Deficiency.
9
+
10
+
11
+ Specifications are based on the decision tree as outlined in Tayoun etal, 2018 (Hum Mutat. 2018 Nov;39(11):1517-1524; PMID: 30192042) SLC6A8: PVS1 is applicable as described in Abou Tayoun et al, 2018 (PMID: 30192042)",None
12
+ SLC6A8 (HGNC:11055),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
13
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
14
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
15
+ SLC6A8 (HGNC:11055),PS1,Strong,PS1 is applicable as described.,None
16
+ SLC6A8 (HGNC:11055),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
17
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
18
+ SLC6A8 (HGNC:11055),PS2,Strong,"Note: Confirmation of paternity in females only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity",None
19
+ SLC6A8 (HGNC:11055),PS2,Moderate,Newly hemizygous male with variant identified in mother.,"Strength,Disease-specific"
20
+ SLC6A8 (HGNC:11055),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
21
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
22
+ SLC6A8 (HGNC:11055),PS3,Strong,"RT-PCR evidence of mis-splicing for non-canonical intronic variants
23
+ For non-canonical splicing variants, RT-PCR evidence demonstrating transcripts of alternative length or specific intron or exon inclusion/exclusion. These studies can be performed in patient derived cells, by placing the mutant genomic DNA into plasmid vectors, or by over-expressing mutant transcript. Assays should demonstrate defective splicing with RT-PCR analysis or RNA sequencing to confirm alternative transcripts.",Disease-specific
24
+ SLC6A8 (HGNC:11055),PS3,Supporting,"Creatine transport activity <10% wild type with less than or equal to 125uM creatine used in SLC6A8 deficient fibroblasts
25
+
26
+
27
+ RT-PCR evidence of mis-splicing for non-canonical intronic variants with evidence of normal splice products.
28
+ For non-canonical splicing variants, RT-PCR evidence demonstrating transcripts of alternative length or specific intron or exon inclusion/exclusion. These studies can be performed in patient derived cells, by placing the mutant genomic DNA into plasmid vectors, or by over-expressing mutant transcript. Assays should demonstrate defective splicing with RT-PCR analysis or RNA sequencing to confirm alternative transcripts.",Disease-specific
29
+ SLC6A8 (HGNC:11055),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
30
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
31
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
32
+ SLC6A8 (HGNC:11055),PS4,Very Strong,"4 independent male probands with elevated urine creatine/creatinine ratio on one occasion at minimum, in addition to any proband used for PP4.
33
+
34
+
35
+ Variant must meet PM2_Supporting criterion for PS4 to apply.",Strength
36
+ SLC6A8 (HGNC:11055),PS4,Strong,"3 independent male probands with elevated urine creatine/creatinine ratio on one occasion at minimum, in addition to any proband used for PP4.
37
+
38
+
39
+ Variant must meet PM2_Supporting criterion for PS4 to apply.",Disease-specific
40
+ SLC6A8 (HGNC:11055),PS4,Moderate,"2 independent male probands with elevated urine creatine/creatinine ratio on one occasion at minimum, in addition to any proband used for PP4.
41
+
42
+
43
+ Variant must meet PM2_Supporting criterion for PS4 to apply.",Strength
44
+ SLC6A8 (HGNC:11055),PS4,Supporting,One independent male proband in addition to any proband used for PP4.,Strength
45
+ SLC6A8 (HGNC:11055),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
46
+ SLC6A8 (HGNC:11055),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
47
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
48
+ SLC6A8 (HGNC:11055),PM2,Supporting,"Absent/rare from controls in an ethnically-matched cohort population sample.
49
+ Threshold: <0.00002 (0.002%) AND 0 hemizygotes in gnomAD.",Disease-specific
50
+ SLC6A8 (HGNC:11055),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
51
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
52
+ SLC6A8 (HGNC:11055),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
53
+ SLC6A8 (HGNC:11055),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,None
54
+ SLC6A8 (HGNC:11055),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
55
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
56
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
57
+ SLC6A8 (HGNC:11055),PM5,Moderate,Missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.,None
58
+ SLC6A8 (HGNC:11055),PM5,Supporting,Missense change at an amino acid residue where a different missense change determined to be likely pathogenic has been seen before.,Strength
59
+ SLC6A8 (HGNC:11055),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
60
+ SLC6A8 (HGNC:11055),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
61
+ Note: May be used as stronger evidence with increasing segregation data.",
62
+ SLC6A8 (HGNC:11055),PP1,Strong,"3 affected segregations + 0 unaffected segregations OR
63
+
64
+
65
+ 2 affected segregations + 3 unaffected segregations",Strength
66
+ SLC6A8 (HGNC:11055),PP1,Moderate,2 affected segregations + 0 unaffected segregations.,Strength
67
+ SLC6A8 (HGNC:11055),PP1,Supporting,1 affected family member + 3 unaffected segregations.,Disease-specific
68
+ SLC6A8 (HGNC:11055),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
69
+ SLC6A8 (HGNC:11055),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
70
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
71
+ SLC6A8 (HGNC:11055),PP3,Supporting,"REVEL score >0.75 for missense variants
72
+
73
+
74
+ In frame deletion or insertion predicted deleterious by PROVEAN, MutPred indel, and MutationTaster.
75
+
76
+
77
+ Predicted impact on splicing by SpliceAI and varSEAK.",None
78
+ SLC6A8 (HGNC:11055),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
79
+ SLC6A8 (HGNC:11055),PP4,Strong,"See points scheme in main document
80
+
81
+
82
+ 4 or more points are required for PP4_Strong; two or more data types must be available to apply PP4_Strong, full sequencing of SLC6A8, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.
83
+
84
+
85
+
86
+
87
+ For PP4, assign points for each type of data present, and then add up the points.
88
+
89
+
90
+
91
+
92
+ PP4 = 1-2 points, PP4_Moderate = 3 points, PP4_Strong = 4 points
93
+
94
+
95
+ Two or more data types are recommended to reach moderate and required to reach strong.
96
+
97
+
98
+ An individual used to assign PP4, at any weight, cannot be also included for PS4 count. If multiple unrelated probands with the variant have been identified, it is recommended that the case with the highest PP4 points is assigned the appropriate weight for PP4, and the other cases are used for PS4.
99
+
100
+
101
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
102
+
103
+
104
+ For PP4 to be applied at strong, full SLC6A8 gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
105
+ SLC6A8 (HGNC:11055),PP4,Moderate,"See points scheme in main document.
106
+
107
+
108
+ 3 points required for PP4_Moderate.
109
+
110
+
111
+
112
+
113
+ For PP4, assign points for each type of data present, and then add up the points.
114
+
115
+
116
+
117
+
118
+ PP4 = 1-2 points, PP4_Moderate = 3 points, PP4_Strong = 4 points
119
+
120
+
121
+ Two or more data types are recommended to reach moderate and required to reach strong.
122
+
123
+
124
+ An individual used to assign PP4, at any weight, cannot be also included for PS4 count. If multiple unrelated probands with the variant have been identified, it is recommended that the case with the highest PP4 points is assigned the appropriate weight for PP4, and the other cases are used for PS4.
125
+
126
+
127
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
128
+
129
+
130
+ For PP4 to be applied at strong, full SLC6A8 gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Strength
131
+ SLC6A8 (HGNC:11055),PP4,Supporting,"Elevated urine creatine/creatinine ratio.
132
+ • 1-2 points required for PP4; at minimum, elevated urine creatine/creatinine level on one occasion.
133
+
134
+
135
+
136
+
137
+ For PP4, assign points for each type of data present, and then add up the points.
138
+
139
+
140
+
141
+
142
+ PP4 = 1-2 points, PP4_Moderate = 3 points, PP4_Strong = 4 points
143
+
144
+
145
+ Two or more data types are recommended to reach moderate and required to reach strong.
146
+
147
+
148
+ An individual used to assign PP4, at any weight, cannot be also included for PS4 count. If multiple unrelated probands with the variant have been identified, it is recommended that the case with the highest PP4 points is assigned the appropriate weight for PP4, and the other cases are used for PS4.
149
+
150
+
151
+ Variant must meet PM2_Supporting for PP4 to apply at any strength.
152
+
153
+
154
+ For PP4 to be applied at strong, full SLC6A8 gene sequencing, including all coding exons and intron/exon boundaries, must have been carried out. If not, consider downgrading.",Disease-specific
155
+ SLC6A8 (HGNC:11055),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
156
+ SLC6A8 (HGNC:11055),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
157
+ SLC6A8 (HGNC:11055),BA1,Stand Alone,Allele frequency >0.0020 (0.2%) OR ≥10 hemizygotes in gnomAD,Disease-specific
158
+ SLC6A8 (HGNC:11055),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
159
+ SLC6A8 (HGNC:11055),BS1,Strong,Allele frequency > 0.0002 (0.02%) OR ≥ 5 hemizygotes in gnomAD,Disease-specific
160
+ SLC6A8 (HGNC:11055),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
161
+ SLC6A8 (HGNC:11055),BS2,Strong,Observed in the homozygous state in a healthy adult,None
162
+ SLC6A8 (HGNC:11055),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
163
+ SLC6A8 (HGNC:11055),BS3,Supporting,"Creatine transport assay demonstrating ≥50% normal transport activity using physiological creatine concentrations (≤125μM creatine).
164
+
165
+
166
+ RT-PCR evidence demonstrating no observable effect of splicing.
167
+
168
+
169
+ Expression assay demonstrating wild type transcript levels","Strength,Disease-specific"
170
+ SLC6A8 (HGNC:11055),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
171
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
172
+ SLC6A8 (HGNC:11055),BS4,Strong,Lack of segregation in affected members of a family.,None
173
+ SLC6A8 (HGNC:11055),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
174
+ SLC6A8 (HGNC:11055),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,NA
175
+ SLC6A8 (HGNC:11055),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
176
+ SLC6A8 (HGNC:11055),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
177
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
178
+ SLC6A8 (HGNC:11055),BP4,Supporting,"REVEL score <0.2 for missense variants
179
+
180
+
181
+ In frame deletion or insertion predicted benign by PROVEAN, MutPred indel, and MutationTaster.
182
+
183
+
184
+ No predicted impact on splicing by SpliceAI and varSEAK.",None
185
+ SLC6A8 (HGNC:11055),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
186
+ SLC6A8 (HGNC:11055),BP5,Supporting,"Variant found in a case with an alternate molecular basis for disease.
187
+ BP5 applicable as described; the case must have specific features of creatine transporter deficiency, such as low creatine on brain magnetic resonance spectroscopy, or elevated urine creatine in order to apply this criterion. Presence of developmental delay and seizures is not sufficient.",None
188
+ SLC6A8 (HGNC:11055),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
189
+ SLC6A8 (HGNC:11055),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
190
+ SLC6A8 (HGNC:11055),BP7,Supporting,A synonymous (silent) variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,None
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCoagulationFactorDeficiencyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforF8Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,122 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ F8 (HGNC:3546),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ F8 (HGNC:3546),PVS1,Very Strong,Per Coagulation Factor Deficiency VCEP/SVI PVS1 decision tree.,General recommendation
9
+ F8 (HGNC:3546),PVS1,Strong,Per Coagulation Factor Deficiency VCEP/SVI PVS1 decision tree.,General recommendation
10
+ F8 (HGNC:3546),PVS1,Moderate,Per Coagulation Factor Deficiency VCEP/SVI PVS1 decision tree.,General recommendation
11
+ F8 (HGNC:3546),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
12
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
13
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
14
+ F8 (HGNC:3546),PS1,Strong,"This evidence code can be applied when there is 1 pathogenic variant or 2 likely pathogenic variants at the same residue based on
15
+ F8
16
+ gene rule specifications from the Coagulation Factor Deficiency VCEP and where
17
+ in silico
18
+ predictors do not suggest a splicing defect.",General recommendation
19
+ F8 (HGNC:3546),PS1,Moderate,"This evidence code can be applied when there is 1 likely pathogenic variants at the same residue based on
20
+ F8
21
+ gene rule specifications from the Coagulation Factor Deficiency VCEP and where
22
+ in silico
23
+ predictors do not suggesting a splicing defect.",General recommendation
24
+ F8 (HGNC:3546),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
25
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
26
+ F8 (HGNC:3546),PS2,Very Strong,Use the SVI recommendations for de novo cases; 4 points. Use de novo guidance below to determine point value.,Disease-specific
27
+ F8 (HGNC:3546),PS2,Strong,Use the SVI recommendations for de novo cases; 2 points. Use de novo guidance below to determine point value.,Disease-specific
28
+ F8 (HGNC:3546),PS2,Moderate,Use the SVI recommendations for de novo cases; 1 point. Use de novo guidance below to determine point value.,Disease-specific
29
+ F8 (HGNC:3546),PS2,Supporting,Use the SVI recommendations for de novo cases; 0.5 point. Use de novo guidance below to determine point value.,Disease-specific
30
+ F8 (HGNC:3546),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
31
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
32
+ F8 (HGNC:3546),PS3,Strong,Abnormal factor VIII activity (<40 IU/dL or 40%) via one-stage or two-stage chromogenic assay in a cell line and/or mouse model.,Disease-specific
33
+ F8 (HGNC:3546),PS3,Supporting,Absent or significantly reduced factor VIII antigen levels compared to wildtype in a cell line by quantitative assay.,Disease-specific
34
+ F8 (HGNC:3546),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
35
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
36
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
37
+ F8 (HGNC:3546),PS4,Very Strong,≥8 probands meet criteria described below,Disease-specific
38
+ F8 (HGNC:3546),PS4,Strong,4-7 probands meet criteria described below,Disease-specific
39
+ F8 (HGNC:3546),PS4,Moderate,2-3 probands meet criteria described below,Disease-specific
40
+ F8 (HGNC:3546),PS4,Supporting,1 proband meets criteria described below,Disease-specific
41
+ F8 (HGNC:3546),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
42
+ F8 (HGNC:3546),PM1,Strong,"This code can be applied at the strong level for variants involving the following residues: R391-S392, R759-S760, E1701-Q1705, R1708-S1709, Y1683, Y1689, Y737, Y742.",Gene-specific
43
+ F8 (HGNC:3546),PM1,Moderate,"This code can be applied at the moderate level for variants involving the following residues: 
44
+
45
+
46
+ Residues affecting secretion: Arg1667, Arg1332
47
+
48
+
49
+ FXa-binding residues: Gly2267-Gly2304 (with the exception of Ser2283)",Gene-specific
50
+ F8 (HGNC:3546),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
51
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
52
+ F8 (HGNC:3546),PM2,Supporting,"Variant must be absent in males in population databases, such as gnomAD.",Disease-specific
53
+ F8 (HGNC:3546),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
54
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
55
+ F8 (HGNC:3546),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
56
+ F8 (HGNC:3546),PM4,Moderate,Use code with no specification.,None
57
+ F8 (HGNC:3546),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
58
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
59
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
60
+ F8 (HGNC:3546),PM5,Moderate,"This evidence code can be applied when there is 1 pathogenic variant or 2 likely pathogenic variants at the same residue based on the
61
+ F8
62
+ rule specifications from the Coagulation Factor Deficiency VCEP and where
63
+ in silico
64
+ predictors do not suggest a splicing defect.",Gene-specific
65
+ F8 (HGNC:3546),PM5,Supporting,"This evidence code can be applied when there is 1 likely pathogenic variant at the same residue based on the
66
+ F8
67
+ rule specifications from the Coagulation Factor Deficiency VCEP and where
68
+ in silico
69
+ predictors do not suggest a splicing defect.",Gene-specific
70
+ F8 (HGNC:3546),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
71
+ F8 (HGNC:3546),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
72
+ Note: May be used as stronger evidence with increasing segregation data.",
73
+ F8 (HGNC:3546),PP1,Strong,The code is application when there are ≥4 meioses across ≥2 families.,Disease-specific
74
+ F8 (HGNC:3546),PP1,Moderate,The code is application when there are at least 3 meioses across one or more families.,Disease-specific
75
+ F8 (HGNC:3546),PP1,Supporting,"The code is application when there are 2 meioses in one family
76
+ OR
77
+ 1 meiosis between 2 affected siblings.",Disease-specific
78
+ F8 (HGNC:3546),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
79
+ F8 (HGNC:3546),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
80
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
81
+ F8 (HGNC:3546),PP3,Supporting,Code can be applied for variants where the REVEL score is greater than or equal to 0.6 or a SpliceAI score of greater than or equal to 0.5.,Gene-specific
82
+ F8 (HGNC:3546),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
83
+ F8 (HGNC:3546),PP4,Moderate,Proband must meet hemophilia A phenotype criteria AND have full gene sequencing and deletion/duplication analysis.,Disease-specific
84
+ F8 (HGNC:3546),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
85
+ F8 (HGNC:3546),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
86
+ F8 (HGNC:3546),BA1,Stand Alone,MAF cutoff of greater or equal to 0.0333% (or 0.000333).,Gene-specific
87
+ F8 (HGNC:3546),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
88
+ F8 (HGNC:3546),BS1,Strong,MAF cutoff of greater than or equal to 0.00333% (or 0.0000333).,Gene-specific
89
+ F8 (HGNC:3546),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
90
+ F8 (HGNC:3546),BS2,Strong,"This evidence code can be used when a 
91
+ F8
92
+ variant is observed in a male with a normal factor VIII activity level (<40% IU) using a one stage and/or a chromogenic assay.",Disease-specific
93
+ F8 (HGNC:3546),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
94
+ F8 (HGNC:3546),BS3,Strong,"This code can be used for 
95
+ F8
96
+ gene variants studied in a cell line or mouse model setting that confer a normal factor VIII activity level AND normal factor VIII antigen level OR normal Western Blot.",Disease-specific
97
+ F8 (HGNC:3546),BS3,Supporting,"This code can be used for 
98
+ F8
99
+ gene variants studied in a cell line or mouse model setting that confer the following results:                    
100
+
101
+
102
+
103
+
104
+ Normal factor VIII activity level, OR
105
+
106
+
107
+ Abnormal factor VIII activity level with abnormal 2N binding assay suggesting a diagnosis of VWD Normandy (VWD 2N) instead of hemophilia A.",Disease-specific
108
+ F8 (HGNC:3546),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
109
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
110
+ F8 (HGNC:3546),BS4,Strong,"This evidence code can be used when a 
111
+ F8
112
+ variant is observed in a male with a family history of hemophilia A and has a normal factor VIII activity level.",Disease-specific
113
+ F8 (HGNC:3546),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
114
+ F8 (HGNC:3546),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,NA
115
+ F8 (HGNC:3546),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
116
+ F8 (HGNC:3546),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
117
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
118
+ F8 (HGNC:3546),BP4,Supporting,This code can be applied for variants reaching a REVEL score of 0.3 or below AND a Splice AI score of less than or equal to 0.05.,Gene-specific
119
+ F8 (HGNC:3546),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
120
+ F8 (HGNC:3546),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
121
+ F8 (HGNC:3546),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
122
+ F8 (HGNC:3546),BP7,Supporting,Splice AI should be used to suggest no splicing impact. Splicing prediction score of less than or equal to 0.05 is required. Conservation should be assess using PhyloP (cutoff less than 0.1) and PhastCons (cutoff less than 0.5).,Gene-specific
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCoagulationFactorDeficiencyExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforF9Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,111 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ F9 (HGNC:3551),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ F9 (HGNC:3551),PVS1,Very Strong,Per Coagulation Factor Deficiency VCEP/SVI PVS1 decision tree.,Gene-specific
9
+ F9 (HGNC:3551),PVS1,Strong,Per Coagulation Factor Deficiency VCEP/SVI PVS1 decision tree.,Gene-specific
10
+ F9 (HGNC:3551),PVS1,Moderate,Per Coagulation Factor Deficiency VCEP/SVI PVS1 decision tree.,Gene-specific
11
+ F9 (HGNC:3551),PVS1,Supporting,Per Coagulation Factor Deficiency VCEP/SVI PVS1 decision tree.,Gene-specific
12
+ F9 (HGNC:3551),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
13
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
14
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
15
+ F9 (HGNC:3551),PS1,Strong,"This evidence code can be applied when there is 1 pathogenic variant or 2 likely pathogenic variants at the same residue based on
16
+ F9
17
+ gene rule specifications from the Coagulation Factor Deficiency VCEP and where
18
+ in silico
19
+ predictors do not suggest a splicing defect.",General recommendation
20
+ F9 (HGNC:3551),PS1,Moderate,"This evidence code can be applied when there is 1 likely pathogenic variants at the same residue based on
21
+ F9
22
+ gene rule specifications from the Coagulation Factor Deficiency VCEP and where
23
+ in silico
24
+ predictors do not suggest a splicing defect.",General recommendation
25
+ F9 (HGNC:3551),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
26
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
27
+ F9 (HGNC:3551),PS2,Very Strong,Use the SVI recommendations for de novo cases; 4 points. Use de novo guidance below to determine point value.,Disease-specific
28
+ F9 (HGNC:3551),PS2,Strong,Use the SVI recommendations for de novo cases; 2 points. Use de novo guidance below to determine point value.,Disease-specific
29
+ F9 (HGNC:3551),PS2,Moderate,Use the SVI recommendations for de novo cases; 1 point. Use de novo guidance below to determine point value.,Disease-specific
30
+ F9 (HGNC:3551),PS2,Supporting,Use the SVI recommendations for de novo cases; 0.5 point. Use de novo guidance below to determine point value.,Disease-specific
31
+ F9 (HGNC:3551),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
32
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
33
+ F9 (HGNC:3551),PS3,Strong,Abnormal factor IX activity level (<40 IU/dL or 40%) in a cell line and/or mouse model.,Disease-specific
34
+ F9 (HGNC:3551),PS3,Moderate,Abnormal factor IX activity level (<40 IU/dL or 40%) studied in an animal model setting other than mouse (i.e. – bovine factor IX activity levels compared to factor X levels).,Disease-specific
35
+ F9 (HGNC:3551),PS3,Supporting,Absent or significantly reduced factor IX antigen level compared to wildtype using conformation-specific reporter assay in cell lines.,Disease-specific
36
+ F9 (HGNC:3551),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
37
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
38
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
39
+ F9 (HGNC:3551),PS4,Very Strong,≥8 probands meet criteria described below,Disease-specific
40
+ F9 (HGNC:3551),PS4,Strong,4-7 probands meet criteria described below,Disease-specific
41
+ F9 (HGNC:3551),PS4,Moderate,2-3 probands meet criteria described below,Disease-specific
42
+ F9 (HGNC:3551),PS4,Supporting,1 proband meets criteria described below,Disease-specific
43
+ F9 (HGNC:3551),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
44
+ F9 (HGNC:3551),PM1,Strong,"This code can be used for variants affecting any of the 3 catalytic residues (H267, D315 or S411) and 2 activation residues (R191-A192 and R226-V227) in the 
45
+ F9
46
+ gene (PMID: 12554099).",Gene-specific
47
+ F9 (HGNC:3551),PM1,Moderate,"This code should be applied when the variant is within exons 3, 4 or 5.",Gene-specific
48
+ F9 (HGNC:3551),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
49
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
50
+ F9 (HGNC:3551),PM2,Supporting,"Variant must be absent in males in population databases, such as gnomAD.",Disease-specific
51
+ F9 (HGNC:3551),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
52
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
53
+ F9 (HGNC:3551),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
54
+ F9 (HGNC:3551),PM4,Moderate,Use code with no specification.,None
55
+ F9 (HGNC:3551),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
56
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
57
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
58
+ F9 (HGNC:3551),PM5,Moderate,"This evidence code can be applied when there is 1 pathogenic variant or 2 likely pathogenic variants at the same residue based on
59
+ F9
60
+ rule specification from the Coagulation Factor Deficiency VCEP and where
61
+ in silico
62
+ predictors do not suggest a splicing defect.",Gene-specific
63
+ F9 (HGNC:3551),PM5,Supporting,"This evidence code can be applied when there is 1 likely pathogenic variant at the same residue based on
64
+ F9
65
+ rule specifications Coagulation Factor Deficiency VCEP and where
66
+ in silico
67
+ predictors do not suggest a splicing defect. A “highly suspicious” VUS is defined as a variant that is 1 supporting code away from reaching a likely pathogenic classification.",Gene-specific
68
+ F9 (HGNC:3551),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
69
+ F9 (HGNC:3551),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
70
+ Note: May be used as stronger evidence with increasing segregation data.",
71
+ F9 (HGNC:3551),PP1,Strong,This code is applicable when there ≥4 meioses across ≥2 families.,Disease-specific
72
+ F9 (HGNC:3551),PP1,Moderate,This code is applicable when there are at least 3 meioses across one or more families.,Disease-specific
73
+ F9 (HGNC:3551),PP1,Supporting,"This code is applicable when there 2 meioses in one family
74
+ OR
75
+ 1 meiosis between 2 affected siblings.",Disease-specific
76
+ F9 (HGNC:3551),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
77
+ F9 (HGNC:3551),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
78
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
79
+ F9 (HGNC:3551),PP3,Supporting,Code can be applied for variants where the REVEL score is greater than or equal to 0.6 or a SpliceAI score of greater than or equal to 0.5.,Gene-specific
80
+ F9 (HGNC:3551),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
81
+ F9 (HGNC:3551),PP4,Moderate,Proband must meet hemophilia B phenotype criteria AND have full gene sequencing and deletion/duplication analysis.,Disease-specific
82
+ F9 (HGNC:3551),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
83
+ F9 (HGNC:3551),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
84
+ F9 (HGNC:3551),BA1,Stand Alone,MAF cutoff of greater than or equal to 0.0000556 (or 0.00556%).,Gene-specific
85
+ F9 (HGNC:3551),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
86
+ F9 (HGNC:3551),BS1,Strong,MAF cutoff of greater than or equal to 0.00000556 (or 0.000556%).,Gene-specific
87
+ F9 (HGNC:3551),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
88
+ F9 (HGNC:3551),BS2,Strong,"This evidence code can be used when a 
89
+ F9
90
+ variant is observed in a male with a normal factor IX activity level (<40% IU).",Disease-specific
91
+ F9 (HGNC:3551),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
92
+ F9 (HGNC:3551),BS3,Strong,"This code can be used for 
93
+ F9
94
+ gene variants studied in a cell line or mouse model setting that confer a normal factor IX activity AND normal factor IX antigen levels 
95
+ OR
96
+ normal Western Blot.",Disease-specific
97
+ F9 (HGNC:3551),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
98
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
99
+ F9 (HGNC:3551),BS4,Strong,"This evidence code can be used when a 
100
+ F9
101
+ variant is observed in a male with a family history of hemophilia B and has a normal factor IX activity level.",Disease-specific
102
+ F9 (HGNC:3551),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
103
+ F9 (HGNC:3551),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,NA
104
+ F9 (HGNC:3551),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
105
+ F9 (HGNC:3551),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
106
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
107
+ F9 (HGNC:3551),BP4,Supporting,This code can be applied for variants reaching a REVEL score of 0.3 or below AND a Splice AI score of less than or equal to 0.01.,Gene-specific
108
+ F9 (HGNC:3551),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
109
+ F9 (HGNC:3551),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
110
+ F9 (HGNC:3551),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
111
+ F9 (HGNC:3551),BP7,Supporting,Splicing prediction score of less than or equal to 0.01 is required. Conservation should be assess using PhyloP (cutoff less than 0.1) and PhastCons (cutoff less than 0.5).,Gene-specific
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACTA1Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,180 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ ACTA1 (HGNC:129),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ ACTA1 (HGNC:129),PVS1,Very Strong,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
9
+ Caveats:
10
+
11
+
12
+
13
+
14
+ Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
15
+
16
+
17
+ Use caution interpreting LOF variants at the extreme 3’ end of a gene.
18
+
19
+
20
+ Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
21
+
22
+
23
+ Use caution in the presence of multiple transcripts.",Gene-specific
24
+ ACTA1 (HGNC:129),PVS1,Strong,See PVS1 flowchart,Gene-specific
25
+ ACTA1 (HGNC:129),PVS1,Moderate,See PVS1 flowchart,Gene-specific
26
+ ACTA1 (HGNC:129),PVS1,Supporting,See PVS1 flowchart,Gene-specific
27
+ ACTA1 (HGNC:129),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
28
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
29
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
30
+ ACTA1 (HGNC:129),PS1,Strong,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
31
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
32
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
33
+ ACTA1 (HGNC:129),PS1,Moderate,No change - use as originally described,No change
34
+ ACTA1 (HGNC:129),PS1,Supporting,No change - use as originally described,No change
35
+ ACTA1 (HGNC:129),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
36
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
37
+ ACTA1 (HGNC:129),PS2,Very Strong,No change - use as originally described,No change
38
+ ACTA1 (HGNC:129),PS2,Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
39
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",No change
40
+ ACTA1 (HGNC:129),PS2,Moderate,No change - use as originally described,No change
41
+ ACTA1 (HGNC:129),PS2,Supporting,No change - use as originally described,No change
42
+ ACTA1 (HGNC:129),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
43
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
44
+ ACTA1 (HGNC:129),PS3,Strong,"Strong may only be considered for variant-specific mouse models. Currently, no other assays are applicable at this strength.",Disease-specific
45
+ ACTA1 (HGNC:129),PS3,Supporting,"There are no specific functional assays listed for the ACTA1 AR specifications; all should be used for AD specifications. However, it is acceptable to use PS3_Supporting for functional analyses if
46
+
47
+
48
+
49
+
50
+ The assay has been validated by a known pathogenic and benign variant AND
51
+
52
+
53
+ There is plausible reason that the function the assay is testing relates to the phenotype AND
54
+
55
+
56
+ The assay conditions are likely to mimic the physiological environment.",Gene-specific
57
+ ACTA1 (HGNC:129),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
58
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
59
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",NA
60
+ ACTA1 (HGNC:129),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
61
+ ACTA1 (HGNC:129),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
62
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
63
+ ACTA1 (HGNC:129),PM2,Supporting,"PM2_Supporting may be applied if the minor allele frequency in population databases of at least 2000 alleles is 
64
+
65
+
66
+ ≤ 0.000005 for autosomal recessive","Disease-specific,Gene-specific"
67
+ ACTA1 (HGNC:129),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
68
+ Note: This requires testing of parents (or offspring) to determine phase.",
69
+ ACTA1 (HGNC:129),PM3,Very Strong,4.0 points per the PM3 chart,Disease-specific
70
+ ACTA1 (HGNC:129),PM3,Strong,2.0 points per the PM3 chart,Gene-specific
71
+ ACTA1 (HGNC:129),PM3,Moderate,"For recessive disorders, detected in trans with a pathogenic variant Note: This requires testing of parents (or offspring) to determine phase.
72
+
73
+
74
+ 1.0 points per the PM3 chart",Gene-specific
75
+ ACTA1 (HGNC:129),PM3,Supporting,0.5 points per the PM3 chart,Gene-specific
76
+ ACTA1 (HGNC:129),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
77
+ ACTA1 (HGNC:129),PM4,Strong,No change - use as originally described,No change
78
+ ACTA1 (HGNC:129),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
79
+ ACTA1 (HGNC:129),PM4,Supporting,No change - use as originally described,No change
80
+ ACTA1 (HGNC:129),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
81
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
82
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
83
+ ACTA1 (HGNC:129),PM5,Strong,No change - use as originally described,No change
84
+ ACTA1 (HGNC:129),PM5,Moderate,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
85
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
86
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
87
+ ACTA1 (HGNC:129),PM5,Supporting,No change - use as originally described,No change
88
+ ACTA1 (HGNC:129),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
89
+ ACTA1 (HGNC:129),PM6,Strong,No change - use as originally described,No change
90
+ ACTA1 (HGNC:129),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity.",No change
91
+ ACTA1 (HGNC:129),PM6,Supporting,No change - use as originally described,No change
92
+ ACTA1 (HGNC:129),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
93
+ Note: May be used as stronger evidence with increasing segregation data.",
94
+ ACTA1 (HGNC:129),PP1,Strong,See segregation chart,General recommendation
95
+ ACTA1 (HGNC:129),PP1,Moderate,See segregation chart,General recommendation
96
+ ACTA1 (HGNC:129),PP1,Supporting,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease. Note: May be used as stronger evidence with increasing segregation data.
97
+
98
+
99
+ See segregation chart",General recommendation
100
+ ACTA1 (HGNC:129),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
101
+ ACTA1 (HGNC:129),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
102
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
103
+ ACTA1 (HGNC:129),PP3,Supporting,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.). Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.
104
+
105
+
106
+ PP3 is met if the REVEL score ≥ 0.7 or if the variant is predicted to impact splicing using SpliceAI score ≥0.5",General recommendation
107
+ ACTA1 (HGNC:129),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
108
+ ACTA1 (HGNC:129),PP4,Strong,"PP4 follows the published SVI guidance from Biesecker et al 2023 (PMID:38103548). For ACTA1, a conservative estimate of the diagnostic yield is 33%, which corresponds to +2 points and a moderate strength in Table 2 of this guidance. However, if the proband meets PP4_Moderate criteria and has had a comprehensive myopathy panel, exome, or genome testing that is negative for all other causes of myopathy, PP4 can be applied at strong, per the SVI guidance. 
109
+
110
+
111
+ The combination of PP1 and PP4 is capped at strong.","Disease-specific,Gene-specific"
112
+ ACTA1 (HGNC:129),PP4,Moderate,"PP4 follows the published SVI guidance from Biesecker et al 2023 (PMID:38103548). For ACTA1, a conservative estimate of the diagnostic yield is 33%, which corresponds to +2 points and a moderate strength in Table 2 of this guidance. 
113
+
114
+
115
+ PP4_Moderate is met with the presence of any of these features
116
+
117
+
118
+ Presence on Muscle Biopsy of:
119
+
120
+
121
+
122
+
123
+ Accumulated thin filaments
124
+
125
+
126
+ Intranuclear rods
127
+
128
+
129
+ Cores, fiber type disproportion
130
+
131
+
132
+ Zebra bodies","Disease-specific,Gene-specific"
133
+ ACTA1 (HGNC:129),PP4,Supporting,"If a biopsy demonstrates a presence of nemaline rods, this is suggestive of ACTA1-related congenital myopathy and can be given PP4 at a supporting level.","Disease-specific,Gene-specific"
134
+ ACTA1 (HGNC:129),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
135
+ ACTA1 (HGNC:129),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
136
+ ACTA1 (HGNC:129),BA1,Stand Alone,"The minor allele frequency using the filtering allele frequency of either exomes or genomes in gnomAD is
137
+ ≥0.0025 for AR variants
138
+ . All continental populations in gnomAD used should have at least 2000 alleles and >1 observation. 
139
+
140
+
141
+ BA1 exclusion variants
142
+ (well-known pathogenic variants that are above the specified BA1 threshold) are as follows: 
143
+
144
+
145
+ NM_001100.4(ACTA1):c.541del (p.Asp181fs)
146
+
147
+
148
+ NM_001100.4(ACTA1):c.121C>T(p.Arg41*)",Gene-specific
149
+ ACTA1 (HGNC:129),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
150
+ ACTA1 (HGNC:129),BS1,Strong,"The minor allele frequency using the filtering allele frequency of either exomes or genomes in gnomAD is
151
+ ≥0.00025 for AR
152
+
153
+ variants
154
+ . All continental populations used in gnomAD should have at least 2000 alleles and >1 observation.",Gene-specific
155
+ ACTA1 (HGNC:129),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
156
+ ACTA1 (HGNC:129),BS2,Strong,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",No change
157
+ ACTA1 (HGNC:129),BS2,Moderate,No change - use as originally described,No change
158
+ ACTA1 (HGNC:129),BS2,Supporting,No change - use as originally described,No change
159
+ ACTA1 (HGNC:129),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,NA
160
+ ACTA1 (HGNC:129),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
161
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
162
+ ACTA1 (HGNC:129),BS4,Strong,"Lack of segregation in affected members of a family.
163
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",No change
164
+ ACTA1 (HGNC:129),BS4,Moderate,No change - use as originally described,No change
165
+ ACTA1 (HGNC:129),BS4,Supporting,No change - use as originally described,No change
166
+ ACTA1 (HGNC:129),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
167
+ ACTA1 (HGNC:129),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
168
+ ACTA1 (HGNC:129),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,No change
169
+ ACTA1 (HGNC:129),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
170
+ ACTA1 (HGNC:129),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
171
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
172
+ ACTA1 (HGNC:129),BP4,Supporting,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc) Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.
173
+
174
+
175
+ BP4 is met if the REVEL score ≤ 0.15 or if the variant is not predicted to impact splicing using SpliceAI.",General recommendation
176
+ ACTA1 (HGNC:129),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
177
+ ACTA1 (HGNC:129),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease.,No change
178
+ ACTA1 (HGNC:129),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
179
+ ACTA1 (HGNC:129),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
180
+ ACTA1 (HGNC:129),BP7,Supporting,A synonymous variant for which SpliceAI predicts no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforACTA1Version2.0.0_version=2.0.0.csv ADDED
@@ -0,0 +1,163 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ ACTA1 (HGNC:129),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",NA
8
+ ACTA1 (HGNC:129),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
9
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
10
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
11
+ ACTA1 (HGNC:129),PS1,Strong,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
12
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
13
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
14
+ ACTA1 (HGNC:129),PS1,Moderate,No change - use as originally described,No change
15
+ ACTA1 (HGNC:129),PS1,Supporting,No change - use as originally described,No change
16
+ ACTA1 (HGNC:129),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
17
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
18
+ ACTA1 (HGNC:129),PS2,Very Strong,No change - use as originally described,No change
19
+ ACTA1 (HGNC:129),PS2,Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
20
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",No change
21
+ ACTA1 (HGNC:129),PS2,Moderate,No change - use as originally described,No change
22
+ ACTA1 (HGNC:129),PS2,Supporting,No change - use as originally described,No change
23
+ ACTA1 (HGNC:129),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
24
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
25
+ ACTA1 (HGNC:129),PS3,Strong,"Strong may only be considered for variant-specific mouse models. Currently, no other assays are applicable at this strength.",Disease-specific
26
+ ACTA1 (HGNC:129),PS3,Moderate,The two assays from PS3_Supporting may be stacked to reach a Moderate Strength,Gene-specific
27
+ ACTA1 (HGNC:129),PS3,Supporting,"Three specific assays are currently suggested to be applied at Supporting:
28
+
29
+
30
+
31
+
32
+ Actin Localization: An abnormal readout consists of integration of actin into cytoplasmic or intranuclear aggregates or rods.
33
+
34
+
35
+ Actin Polymerization: An abnormal readout consists of a significant reduction in levels of actin in insoluble fraction or absent or short polymerized actin filaments compared to WT.
36
+
37
+
38
+ Actin Motility Assay: An abnormal readout consists of percent motility, velocity, and force generation statistically different from WT.
39
+
40
+
41
+
42
+
43
+ If not listed above, it is acceptable to use PS3_Supporting for other functional analyses if
44
+
45
+
46
+
47
+
48
+ The assay has been validated by a known pathogenic and benign variant AND
49
+
50
+
51
+ There is plausible reason that the function the assay is testing relates to the phenotype AND
52
+
53
+
54
+ The assay conditions are likely to mimic the physiological environment.",Gene-specific
55
+ ACTA1 (HGNC:129),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
56
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
57
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
58
+ ACTA1 (HGNC:129),PS4,Strong,8 case observations,Gene-specific
59
+ ACTA1 (HGNC:129),PS4,Moderate,4 case observations,Gene-specific
60
+ ACTA1 (HGNC:129),PS4,Supporting,2 case observations,Gene-specific
61
+ ACTA1 (HGNC:129),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
62
+ ACTA1 (HGNC:129),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
63
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
64
+ ACTA1 (HGNC:129),PM2,Supporting,"PM2_Supporting may be applied if the minor allele frequency in population databases of at least 2000 alleles is 
65
+
66
+
67
+ absent (1 allele allowed) for autosomal dominant","Disease-specific,Gene-specific"
68
+ ACTA1 (HGNC:129),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
69
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
70
+ ACTA1 (HGNC:129),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
71
+ ACTA1 (HGNC:129),PM4,Strong,No change - use as originally described,No change
72
+ ACTA1 (HGNC:129),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
73
+ ACTA1 (HGNC:129),PM4,Supporting,No change - use as originally described,No change
74
+ ACTA1 (HGNC:129),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
75
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
76
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
77
+ ACTA1 (HGNC:129),PM5,Strong,No change - use as originally described,No change
78
+ ACTA1 (HGNC:129),PM5,Moderate,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
79
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
80
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
81
+ ACTA1 (HGNC:129),PM5,Supporting,No change - use as originally described,No change
82
+ ACTA1 (HGNC:129),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
83
+ ACTA1 (HGNC:129),PM6,Strong,No change - use as originally described,No change
84
+ ACTA1 (HGNC:129),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity.",No change
85
+ ACTA1 (HGNC:129),PM6,Supporting,No change - use as originally described,No change
86
+ ACTA1 (HGNC:129),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
87
+ Note: May be used as stronger evidence with increasing segregation data.",
88
+ ACTA1 (HGNC:129),PP1,Strong,See segregation chart,General recommendation
89
+ ACTA1 (HGNC:129),PP1,Moderate,See segregation chart,General recommendation
90
+ ACTA1 (HGNC:129),PP1,Supporting,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease. Note: May be used as stronger evidence with increasing segregation data.
91
+
92
+
93
+ See segregation chart",General recommendation
94
+ ACTA1 (HGNC:129),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,
95
+ ACTA1 (HGNC:129),PP2,Supporting,ACTA1 is a gene that is constrained for missense variation (gnomAD v4.1 z=6.09). PP2 may be used for missense variants with an autosomal dominant mode of inheritance.,Gene-specific
96
+ ACTA1 (HGNC:129),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
97
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
98
+ ACTA1 (HGNC:129),PP3,Supporting,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.). Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.
99
+
100
+
101
+ PP3 is met if the REVEL score ≥ 0.7 or if the variant is predicted to impact splicing using SpliceAI score ≥0.5",General recommendation
102
+ ACTA1 (HGNC:129),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
103
+ ACTA1 (HGNC:129),PP4,Strong,"PP4 follows the published SVI guidance from Biesecker et al 2023 (PMID:38103548). For ACTA1, a conservative estimate of the diagnostic yield is 33%, which corresponds to +2 points and a moderate strength in Table 2 of this guidance. However, if the proband meets PP4_Moderate criteria and has had a comprehensive myopathy panel, exome, or genome testing that is negative for all other causes of myopathy, PP4 can be applied at strong, per the SVI guidance.   
104
+
105
+
106
+ The combination of PP1 and PP4 is capped at strong.","Disease-specific,Gene-specific"
107
+ ACTA1 (HGNC:129),PP4,Moderate,"PP4 follows the published SVI guidance from Biesecker et al 2023 (PMID:38103548). For ACTA1, a conservative estimate of the diagnostic yield is 33%, which corresponds to +2 points and a moderate strength in Table 2 of this guidance. 
108
+
109
+
110
+ PP4_Moderate is met with the presence of any of these features
111
+
112
+
113
+ Presence on Muscle Biopsy of:
114
+
115
+
116
+
117
+
118
+ Accumulated thin filaments
119
+
120
+
121
+ Intranuclear rods
122
+
123
+
124
+ Cores, fiber type disproportion
125
+
126
+
127
+ Zebra bodies","Disease-specific,Gene-specific"
128
+ ACTA1 (HGNC:129),PP4,Supporting,"If a biopsy demonstrates a presence of nemaline rods, this is suggestive of ACTA1-related congenital myopathy and can be given PP4 at a supporting level.","Disease-specific,Gene-specific"
129
+ ACTA1 (HGNC:129),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
130
+ ACTA1 (HGNC:129),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
131
+ ACTA1 (HGNC:129),BA1,Stand Alone,"The minor allele frequency using the filtering allele frequency of either exomes or genomes in gnomAD is
132
+ ≥0.0000781 for AD variants
133
+ . All continental populations in gnomAD used should have at least 2000 alleles and >1 observation.",Gene-specific
134
+ ACTA1 (HGNC:129),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
135
+ ACTA1 (HGNC:129),BS1,Strong,"The minor allele frequency using the filtering allele frequency of either exomes or genomes in gnomAD is
136
+ ≥0.00000781 for AD variants
137
+ . All continental populations used in gnomAD should have at least 2000 alleles and >1 observation.",Gene-specific
138
+ ACTA1 (HGNC:129),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
139
+ ACTA1 (HGNC:129),BS2,Strong,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",No change
140
+ ACTA1 (HGNC:129),BS2,Moderate,No change - use as originally described,No change
141
+ ACTA1 (HGNC:129),BS2,Supporting,No change - use as originally described,No change
142
+ ACTA1 (HGNC:129),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,NA
143
+ ACTA1 (HGNC:129),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
144
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
145
+ ACTA1 (HGNC:129),BS4,Strong,"Lack of segregation in affected members of a family.
146
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",No change
147
+ ACTA1 (HGNC:129),BS4,Moderate,No change - use as originally described,No change
148
+ ACTA1 (HGNC:129),BS4,Supporting,No change - use as originally described,No change
149
+ ACTA1 (HGNC:129),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
150
+ ACTA1 (HGNC:129),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
151
+ ACTA1 (HGNC:129),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,No change
152
+ ACTA1 (HGNC:129),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
153
+ ACTA1 (HGNC:129),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
154
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
155
+ ACTA1 (HGNC:129),BP4,Supporting,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc) Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.
156
+
157
+
158
+ BP4 is met if the REVEL score ≤ 0.15 or if the variant is not predicted to impact splicing using SpliceAI.",General recommendation
159
+ ACTA1 (HGNC:129),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
160
+ ACTA1 (HGNC:129),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease.,No change
161
+ ACTA1 (HGNC:129),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
162
+ ACTA1 (HGNC:129),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
163
+ ACTA1 (HGNC:129),BP7,Supporting,A synonymous variant for which SpliceAI predicts no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDNM2Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,161 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ DNM2 (HGNC:2974),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",NA
8
+ DNM2 (HGNC:2974),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
9
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
10
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
11
+ DNM2 (HGNC:2974),PS1,Strong,Same amino acid change as a previously established pathogenic variant regardless of nucleotide change. Example: Val->Leu caused by either G>C or G>T in the same codon. Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.,No change
12
+ DNM2 (HGNC:2974),PS1,Moderate,No change - use as originally described,No change
13
+ DNM2 (HGNC:2974),PS1,Supporting,No change - use as originally described,No change
14
+ DNM2 (HGNC:2974),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
15
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
16
+ DNM2 (HGNC:2974),PS2,Very Strong,No change - use as originally described,No change
17
+ DNM2 (HGNC:2974),PS2,Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
18
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",No change
19
+ DNM2 (HGNC:2974),PS2,Moderate,No change - use as originally described,No change
20
+ DNM2 (HGNC:2974),PS2,Supporting,No change - use as originally described,No change
21
+ DNM2 (HGNC:2974),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
22
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
23
+ DNM2 (HGNC:2974),PS3,Strong,"Strong may only be considered for variant-specific mouse models. Currently, no other assays are applicable at this strength.",Disease-specific
24
+ DNM2 (HGNC:2974),PS3,Moderate,The two assays from PS3_Supporting may be stacked to reach a Moderate Strength,Gene-specific
25
+ DNM2 (HGNC:2974),PS3,Supporting,"Two specific assays are currently suggested to be applied at Supporting:
26
+
27
+
28
+
29
+
30
+ Oligomerization: An abnormal readout consists of integration of increased dynamin stability compared to WT dynamin.
31
+
32
+
33
+ GTPase activity: An abnormal readout consists of increased GTPase activity or increased stability compared to wild type dynamin.
34
+
35
+
36
+
37
+
38
+ If not listed above, it is acceptable to use PS3_Supporting for other functional analyses if
39
+
40
+
41
+
42
+
43
+ The assay has been validated by a known pathogenic and benign variant AND
44
+
45
+
46
+ There is plausible reason that the function the assay is testing relates to the phenotype AND
47
+
48
+
49
+ The assay conditions are likely to mimic the physiological environment.",Gene-specific
50
+ DNM2 (HGNC:2974),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
51
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
52
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
53
+ DNM2 (HGNC:2974),PS4,Strong,"Specific phenotypes for proband counting include:
54
+
55
+
56
+ A Congenital myopathy panel should be run and negative for other variants (must include BIN1, RYR1, MTM1) AND at least two of these features:
57
+
58
+
59
+
60
+
61
+ Presence on Muscle Biopsy of: Oxidative activity and/or radial stranding with spokes on a wheel appearance with centrally nucleated muscle fibers
62
+
63
+
64
+ Distal weakness
65
+
66
+
67
+ Characteristic muscle imaging (See Figure 9 of Saade et al 2019 PMID: 31060725 for example)
68
+
69
+
70
+ Ophthalmoparesis and Ptosis (both of these must be observed to count this as one phenotype criteria)
71
+
72
+
73
+
74
+
75
+ At least 1 point, please see PS4 chart",Gene-specific
76
+ DNM2 (HGNC:2974),PS4,Moderate,"0.5 points, please see PS4 chart",Gene-specific
77
+ DNM2 (HGNC:2974),PS4,Supporting,"0.25 points, please see PS4 chart",Gene-specific
78
+ DNM2 (HGNC:2974),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
79
+ DNM2 (HGNC:2974),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
80
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
81
+ DNM2 (HGNC:2974),PM2,Supporting,PM2_Supporting may be applied if the minor allele frequency in population databases of at least 2000 alleles is absent (1 allele allowed),"Disease-specific,Gene-specific"
82
+ DNM2 (HGNC:2974),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
83
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
84
+ DNM2 (HGNC:2974),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
85
+ DNM2 (HGNC:2974),PM4,Strong,No change - use as originally described,No change
86
+ DNM2 (HGNC:2974),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
87
+ DNM2 (HGNC:2974),PM4,Supporting,No change - use as originally described,No change
88
+ DNM2 (HGNC:2974),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
89
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
90
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
91
+ DNM2 (HGNC:2974),PM5,Strong,No change - use as originally described,No change
92
+ DNM2 (HGNC:2974),PM5,Moderate,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
93
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
94
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
95
+ DNM2 (HGNC:2974),PM5,Supporting,No change - use as originally described,No change
96
+ DNM2 (HGNC:2974),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
97
+ DNM2 (HGNC:2974),PM6,Strong,No change - use as originally described,No change
98
+ DNM2 (HGNC:2974),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity.",No change
99
+ DNM2 (HGNC:2974),PM6,Supporting,No change - use as originally described,No change
100
+ DNM2 (HGNC:2974),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
101
+ Note: May be used as stronger evidence with increasing segregation data.",
102
+ DNM2 (HGNC:2974),PP1,Strong,At least five segregations,General recommendation
103
+ DNM2 (HGNC:2974),PP1,Moderate,Four segregations,General recommendation
104
+ DNM2 (HGNC:2974),PP1,Supporting,At least 2 segregations,General recommendation
105
+ DNM2 (HGNC:2974),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,
106
+ DNM2 (HGNC:2974),PP2,Supporting,DNM2 is a gene that is constrained for missense variation (gnomAD v4.1 z=4.87). PP2 may be used for missense variants.,Gene-specific
107
+ DNM2 (HGNC:2974),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
108
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
109
+ DNM2 (HGNC:2974),PP3,Supporting,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.). Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.
110
+
111
+
112
+ PP3 is met if the REVEL score ≥ 0.7 or if the variant is predicted to impact splicing using SpliceAI score ≥0.5",General recommendation
113
+ DNM2 (HGNC:2974),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
114
+ DNM2 (HGNC:2974),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
115
+ DNM2 (HGNC:2974),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
116
+ DNM2 (HGNC:2974),BA1,Stand Alone,"The minor allele frequency using the filtering allele frequency of either exomes or genomes in gnomAD is
117
+ ≥0.0000015
118
+ . All continental populations used in gnomAD should have at least 2000 alleles and >1 observation.","Disease-specific,Gene-specific"
119
+ DNM2 (HGNC:2974),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
120
+ DNM2 (HGNC:2974),BS1,Strong,"The minor allele frequency using the filtering allele frequency of either exomes or genomes in gnomAD is
121
+ ≥0.00000015
122
+ . All continental populations used in gnomAD should have at least 2000 alleles and >1 observation.","Disease-specific,Gene-specific"
123
+ DNM2 (HGNC:2974),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
124
+ DNM2 (HGNC:2974),BS2,Strong,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",No change
125
+ DNM2 (HGNC:2974),BS2,Moderate,No change - use as originally described,No change
126
+ DNM2 (HGNC:2974),BS2,Supporting,No change - use as originally described,No change
127
+ DNM2 (HGNC:2974),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
128
+ DNM2 (HGNC:2974),BS3,Supporting,"BS3_Supporting is met if
129
+ both
130
+ of these assays have WT readouts:
131
+
132
+
133
+ Oligomerization Assay: DNM2 assembly/disassembly dynamics similar to wild type DNM2
134
+
135
+
136
+ GTPase Activity Assay: GTPase activity and stability similar to wild type DNM2",Gene-specific
137
+ DNM2 (HGNC:2974),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
138
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
139
+ DNM2 (HGNC:2974),BS4,Strong,"Lack of segregation in affected members of a family.
140
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",No change
141
+ DNM2 (HGNC:2974),BS4,Moderate,No change - use as originally described,No change
142
+ DNM2 (HGNC:2974),BS4,Supporting,No change - use as originally described,No change
143
+ DNM2 (HGNC:2974),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
144
+ DNM2 (HGNC:2974),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
145
+ DNM2 (HGNC:2974),BP2,Strong,No change - use as originally described,No change
146
+ DNM2 (HGNC:2974),BP2,Moderate,No change - use as originally described,No change
147
+ DNM2 (HGNC:2974),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,No change
148
+ DNM2 (HGNC:2974),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
149
+ DNM2 (HGNC:2974),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
150
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
151
+ DNM2 (HGNC:2974),BP4,Supporting,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc) Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.
152
+
153
+
154
+ BP4 is met if the REVEL score ≤ 0.15 or if the variant is not predicted to impact splicing using SpliceAI.",General recommendation
155
+ DNM2 (HGNC:2974),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
156
+ DNM2 (HGNC:2974),BP5,Strong,No change - use as originally described,No change
157
+ DNM2 (HGNC:2974),BP5,Moderate,No change - use as originally described,No change
158
+ DNM2 (HGNC:2974),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease.,No change
159
+ DNM2 (HGNC:2974),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
160
+ DNM2 (HGNC:2974),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
161
+ DNM2 (HGNC:2974),BP7,Supporting,A synonymous variant for which SpliceAI predicts no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforMTM1Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,182 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ MTM1 (HGNC:7448),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ MTM1 (HGNC:7448),PVS1,Very Strong,See PVS1 flowchart,Gene-specific
9
+ MTM1 (HGNC:7448),PVS1,Strong,See PVS1 flowchart,Gene-specific
10
+ MTM1 (HGNC:7448),PVS1,Moderate,See PVS1 flowchart,Gene-specific
11
+ MTM1 (HGNC:7448),PVS1,Supporting,See PVS1 flowchart,Gene-specific
12
+ MTM1 (HGNC:7448),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
13
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
14
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
15
+ MTM1 (HGNC:7448),PS1,Strong,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
16
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
17
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
18
+ MTM1 (HGNC:7448),PS1,Moderate,No change - use as originally described,No change
19
+ MTM1 (HGNC:7448),PS1,Supporting,No change - use as originally described,No change
20
+ MTM1 (HGNC:7448),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
21
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
22
+ MTM1 (HGNC:7448),PS2,Very Strong,No change - use as originally described,No change
23
+ MTM1 (HGNC:7448),PS2,Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
24
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",No change
25
+ MTM1 (HGNC:7448),PS2,Moderate,No change - use as originally described,No change
26
+ MTM1 (HGNC:7448),PS2,Supporting,No change - use as originally described,No change
27
+ MTM1 (HGNC:7448),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
28
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
29
+ MTM1 (HGNC:7448),PS3,Strong,"Strong may only be considered for variant-specific mouse models. Currently, no other assays are applicable at this strength.",Disease-specific
30
+ MTM1 (HGNC:7448),PS3,Moderate,"Where indicated, some of the assays from PS3_Supporting may be stacked to reach a Moderate Strength",Gene-specific
31
+ MTM1 (HGNC:7448),PS3,Supporting,"Five specific assays are currently suggested to be applied at Supporting:
32
+
33
+
34
+
35
+
36
+ Phosphatase Activity: An abnormal readout consists of reduced phosphatase activity (measured via levels of PtdIns and PtdIns5p or increased levels of precursors)
37
+
38
+
39
+ Myotubularin Localization
40
+ **
41
+ : An abnormal readout consists of altered localization (presence in spots/aggregates/extensions, loss of cytoplasmic localization)
42
+
43
+
44
+ Myotubularin Translocation
45
+ **
46
+ : An abnormal readout consists of loss of MTM1 recruitment to late endosomal compartment following EGF stimulation
47
+
48
+
49
+ Intracellular Trafficking (Trafficking of endosomal cargo is thought to require phosphoinositide conversion and may be disrupted by defective MTM1 activity): An abnormal readout consists of reduced localization/trafficking of receptor proteins
50
+
51
+
52
+ Protein and Lipid Association: An abnormal readout consists of decreased association with known binding partner (BIN1, Desmin, hVPS34-PI 3-Kinase, MTMR12, Phosphoinositide, or EEA Membrane association)
53
+
54
+
55
+
56
+
57
+ **These two assays should not be stacked because they may not be assessing independent biological function
58
+
59
+
60
+ If not listed above, it is acceptable to use PS3_Supporting for other functional analyses if
61
+
62
+
63
+
64
+
65
+ The assay has been validated by a known pathogenic and benign variant AND
66
+
67
+
68
+ There is plausible reason that the function the assay is testing relates to the phenotype AND
69
+
70
+
71
+ The assay conditions are likely to mimic the physiological environment.",Gene-specific
72
+ MTM1 (HGNC:7448),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
73
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
74
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
75
+ MTM1 (HGNC:7448),PS4,Strong,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls. Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance. Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.
76
+
77
+
78
+ In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.
79
+
80
+
81
+ 5+ cases","Disease-specific,Gene-specific"
82
+ MTM1 (HGNC:7448),PS4,Moderate,3-4 cases,"Disease-specific,Gene-specific"
83
+ MTM1 (HGNC:7448),PS4,Supporting,2 cases,"Disease-specific,Gene-specific"
84
+ MTM1 (HGNC:7448),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
85
+ MTM1 (HGNC:7448),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
86
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
87
+ MTM1 (HGNC:7448),PM2,Supporting,PM2_Supporting may be applied if the minor allele frequency in population databases of at least 2000 alleles is absent (1 observation allowed in females only),"Disease-specific,Gene-specific"
88
+ MTM1 (HGNC:7448),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
89
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
90
+ MTM1 (HGNC:7448),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
91
+ MTM1 (HGNC:7448),PM4,Strong,No change - use as originally described,No change
92
+ MTM1 (HGNC:7448),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
93
+ MTM1 (HGNC:7448),PM4,Supporting,No change - use as originally described,No change
94
+ MTM1 (HGNC:7448),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
95
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
96
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
97
+ MTM1 (HGNC:7448),PM5,Strong,No change - use as originally described,No change
98
+ MTM1 (HGNC:7448),PM5,Moderate,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
99
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
100
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
101
+ MTM1 (HGNC:7448),PM5,Supporting,No change - use as originally described,No change
102
+ MTM1 (HGNC:7448),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
103
+ MTM1 (HGNC:7448),PM6,Strong,No change - use as originally described,No change
104
+ MTM1 (HGNC:7448),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity.",No change
105
+ MTM1 (HGNC:7448),PM6,Supporting,No change - use as originally described,No change
106
+ MTM1 (HGNC:7448),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
107
+ Note: May be used as stronger evidence with increasing segregation data.",
108
+ MTM1 (HGNC:7448),PP1,Strong,At least five segregations,General recommendation
109
+ MTM1 (HGNC:7448),PP1,Moderate,3-4 segregations,General recommendation
110
+ MTM1 (HGNC:7448),PP1,Supporting,2 segregations,General recommendation
111
+ MTM1 (HGNC:7448),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
112
+ MTM1 (HGNC:7448),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
113
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
114
+ MTM1 (HGNC:7448),PP3,Supporting,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.). Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.
115
+
116
+
117
+ PP3 is met if the REVEL score ≥ 0.7 or if the variant is predicted to impact splicing using SpliceAI score ≥0.5",General recommendation
118
+ MTM1 (HGNC:7448),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
119
+ MTM1 (HGNC:7448),PP4,Supporting,"Affected Males (must be negative for BIN1, RYR1, and DNM2 variants)
120
+
121
+
122
+
123
+
124
+ Muscle biopsy with rounded muscle fibers with a single centrally located nucleus surrounded by a halo devoid of contractile elements, but containing mitochondria
125
+
126
+
127
+
128
+
129
+ Carrier Females (a panel test for neuromuscular disease to rule out other causes)
130
+
131
+
132
+
133
+
134
+ Observation of myopathy (may be asymmetric) AND at least 1 other feature
135
+
136
+
137
+ Unilateral skeletal asymmetry
138
+
139
+
140
+ Narrow, elongated face
141
+
142
+
143
+ High arched palate","Disease-specific,Gene-specific"
144
+ MTM1 (HGNC:7448),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
145
+ MTM1 (HGNC:7448),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
146
+ MTM1 (HGNC:7448),BA1,Stand Alone,"The minor allele frequency using the filtering allele frequency of either exomes or genomes in gnomAD is
147
+ ≥0.000016
148
+ . All continental gnomAD populations used should have at least 2000 alleles and >1 observation.","Disease-specific,Gene-specific"
149
+ MTM1 (HGNC:7448),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
150
+ MTM1 (HGNC:7448),BS1,Strong,"The minor allele frequency using the filtering allele frequency of either exomes or genomes in gnomAD is
151
+ ≥0.0000016
152
+ . All continental gnomAD populations used should have at least 2000 alleles and >1 observation.","Disease-specific,Gene-specific"
153
+ MTM1 (HGNC:7448),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
154
+ MTM1 (HGNC:7448),BS2,Strong,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",No change
155
+ MTM1 (HGNC:7448),BS2,Moderate,No change - use as originally described,No change
156
+ MTM1 (HGNC:7448),BS2,Supporting,No change - use as originally described,No change
157
+ MTM1 (HGNC:7448),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,NA
158
+ MTM1 (HGNC:7448),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
159
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
160
+ MTM1 (HGNC:7448),BS4,Strong,"Lack of segregation in affected members of a family.
161
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",No change
162
+ MTM1 (HGNC:7448),BS4,Moderate,No change - use as originally described,No change
163
+ MTM1 (HGNC:7448),BS4,Supporting,No change - use as originally described,No change
164
+ MTM1 (HGNC:7448),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
165
+ MTM1 (HGNC:7448),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
166
+ MTM1 (HGNC:7448),BP2,Strong,No change - use as originally described,No change
167
+ MTM1 (HGNC:7448),BP2,Moderate,No change - use as originally described,No change
168
+ MTM1 (HGNC:7448),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,No change
169
+ MTM1 (HGNC:7448),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
170
+ MTM1 (HGNC:7448),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
171
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
172
+ MTM1 (HGNC:7448),BP4,Supporting,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc) Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.
173
+
174
+
175
+ BP4 is met if the REVEL score ≤ 0.15 or if the variant is not predicted to impact splicing using SpliceAI.",General recommendation
176
+ MTM1 (HGNC:7448),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
177
+ MTM1 (HGNC:7448),BP5,Strong,No change - use as originally described,No change
178
+ MTM1 (HGNC:7448),BP5,Moderate,No change - use as originally described,No change
179
+ MTM1 (HGNC:7448),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease.,No change
180
+ MTM1 (HGNC:7448),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
181
+ MTM1 (HGNC:7448),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
182
+ MTM1 (HGNC:7448),BP7,Supporting,A synonymous variant for which SpliceAI predicts no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforNEBVersion1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,218 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ NEB (HGNC:7720),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ NEB (HGNC:7720),PVS1,Very Strong,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Caveats:
9
+
10
+
11
+
12
+
13
+ Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
14
+
15
+
16
+ Use caution interpreting LOF variants at the extreme 3’ end of a gene.
17
+
18
+
19
+ Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
20
+
21
+
22
+ Use caution in the presence of multiple transcripts.
23
+
24
+
25
+
26
+
27
+ Specified critical regions in NEB that should also receive PVS1 are listed below and in the PVS1 flowchart:
28
+
29
+
30
+ In-frame deletions due to the repetitive nature of NEB, particularly in exon 55, are deleterious and pathogenic (Anderson 2004 PMID:15221447, Lehtokari 2009 PMID:19232495). 
31
+
32
+
33
+ Exons 161-183 (Pelin 1999 PMID:10051637).",Gene-specific
34
+ NEB (HGNC:7720),PVS1,Strong,"In-frame deletions due to the repetitive nature of NEB, particularly in exon 55, are deleterious and pathogenic (Anderson 2004 PMID:15221447, Lehtokari 2009 PMID:19232495). The majority of NEB exons are in frame (exons 3-180, 182); thus, skipping of in-frame exons should be scored at PVS1_Strong.",Gene-specific
35
+ NEB (HGNC:7720),PVS1,Moderate,See PVS1 flowchart,Gene-specific
36
+ NEB (HGNC:7720),PVS1,Supporting,See PVS1 flowchart,Gene-specific
37
+ NEB (HGNC:7720),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
38
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
39
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
40
+ NEB (HGNC:7720),PS1,Strong,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
41
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
42
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
43
+ NEB (HGNC:7720),PS1,Moderate,No change - use as originally described,No change
44
+ NEB (HGNC:7720),PS1,Supporting,No change - use as originally described,No change
45
+ NEB (HGNC:7720),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
46
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
47
+ NEB (HGNC:7720),PS2,Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
48
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",No change
49
+ NEB (HGNC:7720),PS2,Moderate,No change - use as originally described,No change
50
+ NEB (HGNC:7720),PS2,Supporting,No change - use as originally described,No change
51
+ NEB (HGNC:7720),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
52
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
53
+ NEB (HGNC:7720),PS3,Strong,"Strong may only be considered for variant-specific mouse models. Currently, no other assays are applicable at this strength.",Disease-specific
54
+ NEB (HGNC:7720),PS3,Moderate,The two assays from PS3_Supporting may be stacked to reach a Moderate Strength,Gene-specific
55
+ NEB (HGNC:7720),PS3,Supporting,"Two specific assays are currently suggested to be applied at Supporting:
56
+
57
+
58
+
59
+
60
+ Thin filament structure: An abnormal readout consists of a significant difference in intensity reflections of X-ray diffraction patterns generated by muscle fibers from patient biopsies compared to WT.
61
+
62
+
63
+ In vitro
64
+ motility: An abnormal readout consists of a significant difference in speed of single fibers derived from patient muscle compared to WT.
65
+
66
+
67
+
68
+
69
+ If not listed above, it is acceptable to use PS3_Supporting for other functional analyses if
70
+
71
+
72
+
73
+
74
+ The assay has been validated by a known pathogenic and benign variant AND
75
+
76
+
77
+ There is plausible reason that the function the assay is testing relates to the phenotype AND
78
+
79
+
80
+ The assay conditions are likely to mimic the physiological environment.",Gene-specific
81
+ NEB (HGNC:7720),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
82
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
83
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
84
+ NEB (HGNC:7720),PS4,Strong,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls. Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance. 
85
+
86
+
87
+ NEB is associated with Autosomal Recessive disease; PS4 can only be used for case-control studies and not proband counting. Please use PM3 for individual case observations.","Disease-specific,Gene-specific"
88
+ NEB (HGNC:7720),PS4,Moderate,NEB is associated with Autosomal Recessive disease; PS4 can only be used for case-control studies and not proband counting. Please use PM3 for individual case observations.,Gene-specific
89
+ NEB (HGNC:7720),PS4,Supporting,NEB is associated with Autosomal Recessive disease; PS4 can only be used for case-control studies and not proband counting. Please use PM3 for individual case observations.,Gene-specific
90
+ NEB (HGNC:7720),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
91
+ NEB (HGNC:7720),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
92
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
93
+ NEB (HGNC:7720),PM2,Supporting,PM2_Supporting may be applied if the minor allele frequency in population databases of at least 2000 alleles is ≤ 0.0000559. 1 allele is allowed.,"Disease-specific,Gene-specific"
94
+ NEB (HGNC:7720),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
95
+ Note: This requires testing of parents (or offspring) to determine phase.",
96
+ NEB (HGNC:7720),PM3,Very Strong,4.0 points per the PM3 chart,Disease-specific
97
+ NEB (HGNC:7720),PM3,Strong,2.0 points per the PM3 chart,Disease-specific
98
+ NEB (HGNC:7720),PM3,Moderate,"For recessive disorders, detected in trans with a pathogenic variant Note: This requires testing of parents (or offspring) to determine phase.
99
+
100
+
101
+ 1.0 points per the PM3 chart",Disease-specific
102
+ NEB (HGNC:7720),PM3,Supporting,0.5 points per the PM3 chart,Disease-specific
103
+ NEB (HGNC:7720),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
104
+ NEB (HGNC:7720),PM4,Strong,"In-frame deletions due to the repetitive nature of NEB, particularly in exon 55, are deleterious and pathogenic (Anderson 2004 PMID:15221447, Lehtokari 2009 PMID:19232495).",Gene-specific
105
+ NEB (HGNC:7720),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
106
+ NEB (HGNC:7720),PM4,Supporting,No change - use as originally described,No change
107
+ NEB (HGNC:7720),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
108
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
109
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
110
+ NEB (HGNC:7720),PM5,Strong,No change - use as originally described,No change
111
+ NEB (HGNC:7720),PM5,Moderate,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
112
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
113
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
114
+ NEB (HGNC:7720),PM5,Supporting,No change - use as originally described,No change
115
+ NEB (HGNC:7720),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
116
+ NEB (HGNC:7720),PM6,Strong,No change - use as originally described,No change
117
+ NEB (HGNC:7720),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity.",No change
118
+ NEB (HGNC:7720),PM6,Supporting,No change - use as originally described,No change
119
+ NEB (HGNC:7720),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
120
+ Note: May be used as stronger evidence with increasing segregation data.",
121
+ NEB (HGNC:7720),PP1,Strong,See segregation chart,General recommendation
122
+ NEB (HGNC:7720),PP1,Moderate,See segregation chart,General recommendation
123
+ NEB (HGNC:7720),PP1,Supporting,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease. Note: May be used as stronger evidence with increasing segregation data.
124
+
125
+
126
+ See segregation chart",General recommendation
127
+ NEB (HGNC:7720),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
128
+ NEB (HGNC:7720),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
129
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
130
+ NEB (HGNC:7720),PP3,Supporting,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.). Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.
131
+
132
+
133
+ PP3 is met if the REVEL score ≥ 0.7 or if the variant is predicted to impact splicing using SpliceAI score ≥0.5",General recommendation
134
+ NEB (HGNC:7720),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
135
+ NEB (HGNC:7720),PP4,Supporting,"PP4 is met with the presence of any of these features
136
+
137
+
138
+ Presence on Muscle Biopsy of:
139
+
140
+
141
+
142
+
143
+ Nemaline rods
144
+
145
+
146
+ Shorter thin filaments
147
+
148
+
149
+
150
+
151
+ Functional assays performed upon patient muscle biopsy indicate:
152
+
153
+
154
+
155
+
156
+ Significantly altered Calcium sensitivity
157
+
158
+
159
+ Significantly altered muscle mechanics (altered force-sarcomere length or reduced contractile strength and force generation)","Disease-specific,Gene-specific"
160
+ NEB (HGNC:7720),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
161
+ NEB (HGNC:7720),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
162
+ NEB (HGNC:7720),BA1,Stand Alone,"The minor allele frequency using the filtering allele frequency of either exomes or genomes in gnomAD is ≥0.00559. All populations used should have at least 2000 alleles and >1 observation. The Ashkenazi Jewish, European Finnish, and Other populations in gnomAD will not be used for BA1 application.
163
+
164
+
165
+ BA1 exclusion variants
166
+ (well-known pathogenic variants that are above the specified BA1 threshold) are as follows: 
167
+
168
+
169
+ Exon 55 deletion common in the AJ population (NM_001271208.2:c.7431+1919_7536+374del)
170
+
171
+
172
+ NM_001271208.2:c.19097G>T (p.Ser6366Ile)
173
+
174
+
175
+ NM_001271208.2:c.22249A>C (p.Thr7417Pro)","Disease-specific,Gene-specific"
176
+ NEB (HGNC:7720),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
177
+ NEB (HGNC:7720),BS1,Strong,The minor allele frequency using the filtering allele frequency of either exomes or genomes in gnomAD is ≥0.000237 in continental populations. All populations used should have at least 2000 alleles and >1 observation.,"Disease-specific,Gene-specific"
178
+ NEB (HGNC:7720),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
179
+ NEB (HGNC:7720),BS2,Strong,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",No change
180
+ NEB (HGNC:7720),BS2,Moderate,No change - use as originally described,No change
181
+ NEB (HGNC:7720),BS2,Supporting,No change - use as originally described,No change
182
+ NEB (HGNC:7720),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
183
+ NEB (HGNC:7720),BS3,Moderate,The two assays from BS3_Supporting may be stacked to reach a Moderate Strength,Gene-specific
184
+ NEB (HGNC:7720),BS3,Supporting,"Two specific assays are currently suggested to be applied at Supporting:
185
+
186
+
187
+
188
+
189
+ Thin filament structure: A normal readout consists of no significant difference in intensity reflections of X-ray diffraction patterns generated by muscle fibers from patient biopsies compared to WT.
190
+
191
+
192
+ In vitro
193
+ motility: A normal readout consists of no a significant difference in speed of single fibers derived from patient muscle compared to WT.",Gene-specific
194
+ NEB (HGNC:7720),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
195
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
196
+ NEB (HGNC:7720),BS4,Strong,"Lack of segregation in affected members of a family.
197
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",No change
198
+ NEB (HGNC:7720),BS4,Moderate,No change - use as originally described,No change
199
+ NEB (HGNC:7720),BS4,Supporting,No change - use as originally described,No change
200
+ NEB (HGNC:7720),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
201
+ NEB (HGNC:7720),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
202
+ NEB (HGNC:7720),BP2,Strong,No change - use as originally described,No change
203
+ NEB (HGNC:7720),BP2,Moderate,No change - use as originally described,No change
204
+ NEB (HGNC:7720),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,No change
205
+ NEB (HGNC:7720),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
206
+ NEB (HGNC:7720),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
207
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
208
+ NEB (HGNC:7720),BP4,Supporting,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc) Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.
209
+
210
+
211
+ BP4 is met if the REVEL score ≤ 0.15 or if the variant is not predicted to impact splicing using SpliceAI.",General recommendation
212
+ NEB (HGNC:7720),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
213
+ NEB (HGNC:7720),BP5,Strong,No change - use as originally described,No change
214
+ NEB (HGNC:7720),BP5,Moderate,No change - use as originally described,No change
215
+ NEB (HGNC:7720),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease.,No change
216
+ NEB (HGNC:7720),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
217
+ NEB (HGNC:7720),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
218
+ NEB (HGNC:7720),BP7,Supporting,A synonymous variant for which SpliceAI predicts no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRYR1Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,187 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ RYR1 (HGNC:10483),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ RYR1 (HGNC:10483),PVS1,Very Strong,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
9
+ Caveats:
10
+
11
+
12
+
13
+
14
+ Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
15
+
16
+
17
+ Use caution interpreting LOF variants at the extreme 3’ end of a gene.
18
+
19
+
20
+ Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
21
+
22
+
23
+ Use caution in the presence of multiple transcripts.","Disease-specific,Gene-specific"
24
+ RYR1 (HGNC:10483),PVS1,Strong,"In-frame deletions or in frame exon-skipping variants in the pore/transmembrane region of RYR1 should be scored at PVS1_strong. (Amino acids 4800-4950, Exons 100-103)",Gene-specific
25
+ RYR1 (HGNC:10483),PVS1,Moderate,See PVS1 flowchart,"Disease-specific,Gene-specific"
26
+ RYR1 (HGNC:10483),PVS1,Supporting,See PVS1 flowchart,"Disease-specific,Gene-specific"
27
+ RYR1 (HGNC:10483),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
28
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
29
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
30
+ RYR1 (HGNC:10483),PS1,Strong,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
31
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
32
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
33
+ RYR1 (HGNC:10483),PS1,Moderate,No change - use as originally described,No change
34
+ RYR1 (HGNC:10483),PS1,Supporting,No change - use as originally described,No change
35
+ RYR1 (HGNC:10483),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
36
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
37
+ RYR1 (HGNC:10483),PS2,Very Strong,No change - use as originally described,No change
38
+ RYR1 (HGNC:10483),PS2,Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
39
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",No change
40
+ RYR1 (HGNC:10483),PS2,Moderate,No change - use as originally described,No change
41
+ RYR1 (HGNC:10483),PS2,Supporting,No change - use as originally described,No change
42
+ RYR1 (HGNC:10483),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
43
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
44
+ RYR1 (HGNC:10483),PS3,Strong,"Strong may only be considered for variant-specific mouse models. Currently, no other assays are applicable at this strength.",Disease-specific
45
+ RYR1 (HGNC:10483),PS3,Supporting,"There are no specified functional assay for AR RYR1-related myopathy. However, it is acceptable to use PS3_Supporting for other functional analyses if
46
+
47
+
48
+
49
+
50
+ The assay has been validated by a known pathogenic and benign variant AND
51
+
52
+
53
+ There is plausible reason that the function the assay is testing relates to the phenotype AND
54
+
55
+
56
+ The assay conditions are likely to mimic the physiological environment.",Gene-specific
57
+ RYR1 (HGNC:10483),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
58
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
59
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",NA
60
+ RYR1 (HGNC:10483),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
61
+ RYR1 (HGNC:10483),PM1,Moderate,The pore/transmembrane region of RYR1 is critical for protein function and missense variants that fall within this region (amino acids 4800-4950),"Disease-specific,Gene-specific"
62
+ RYR1 (HGNC:10483),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
63
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
64
+ RYR1 (HGNC:10483),PM2,Supporting,"PM2_Supporting may be applied if the minor allele frequency in population databases of at least 2000 alleles is 
65
+
66
+
67
+  ≤ 0.00000697 for autosomal recessive","Disease-specific,Gene-specific"
68
+ RYR1 (HGNC:10483),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
69
+ Note: This requires testing of parents (or offspring) to determine phase.",
70
+ RYR1 (HGNC:10483),PM3,Very Strong,4.0 points per the PM3 chart,"Disease-specific,Gene-specific"
71
+ RYR1 (HGNC:10483),PM3,Strong,2.0 points per the PM3 chart,"Disease-specific,Gene-specific"
72
+ RYR1 (HGNC:10483),PM3,Moderate,"For recessive disorders, detected in trans with a pathogenic variant Note: This requires testing of parents (or offspring) to determine phase.
73
+
74
+
75
+ 1.0 points per the PM3 chart","Disease-specific,Gene-specific"
76
+ RYR1 (HGNC:10483),PM3,Supporting,0.5 points per the PM3 chart,"Disease-specific,Gene-specific"
77
+ RYR1 (HGNC:10483),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
78
+ RYR1 (HGNC:10483),PM4,Strong,No change - use as originally described,No change
79
+ RYR1 (HGNC:10483),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
80
+ RYR1 (HGNC:10483),PM4,Supporting,No change - use as originally described,No change
81
+ RYR1 (HGNC:10483),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
82
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
83
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
84
+ RYR1 (HGNC:10483),PM5,Strong,No change - use as originally described,No change
85
+ RYR1 (HGNC:10483),PM5,Moderate,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
86
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
87
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
88
+ RYR1 (HGNC:10483),PM5,Supporting,No change - use as originally described,No change
89
+ RYR1 (HGNC:10483),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
90
+ RYR1 (HGNC:10483),PM6,Strong,No change - use as originally described,No change
91
+ RYR1 (HGNC:10483),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity.",No change
92
+ RYR1 (HGNC:10483),PM6,Supporting,No change - use as originally described,No change
93
+ RYR1 (HGNC:10483),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
94
+ Note: May be used as stronger evidence with increasing segregation data.",
95
+ RYR1 (HGNC:10483),PP1,Strong,See segregation chart,General recommendation
96
+ RYR1 (HGNC:10483),PP1,Moderate,See segregation chart,General recommendation
97
+ RYR1 (HGNC:10483),PP1,Supporting,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease. Note: May be used as stronger evidence with increasing segregation data.
98
+
99
+
100
+ See segregation chart",General recommendation
101
+ RYR1 (HGNC:10483),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
102
+ RYR1 (HGNC:10483),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
103
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
104
+ RYR1 (HGNC:10483),PP3,Supporting,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.). Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.
105
+
106
+
107
+ PP3 is met if the REVEL score ≥ 0.7 or if the variant is predicted to impact splicing using SpliceAI score ≥0.5",General recommendation
108
+ RYR1 (HGNC:10483),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
109
+ RYR1 (HGNC:10483),PP4,Supporting,"To meet PP4, a congenital myopathy testing panel should be performed without identification of other causative variants and AT LEAST TWO of these features should be present
110
+
111
+
112
+
113
+
114
+ Presence on Muscle Biopsy of:
115
+ mini cores or central cores (histology or electron microscopy)
116
+
117
+
118
+ Exercise, heat, or anesthetic induced rhabdomyolysis
119
+
120
+
121
+ Ophthalmoplegia
122
+
123
+
124
+ Characteristic muscle imaging (see Figure 8, Saade et al 2019 PMID: 31060725)","Disease-specific,Gene-specific"
125
+ RYR1 (HGNC:10483),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
126
+ RYR1 (HGNC:10483),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
127
+ RYR1 (HGNC:10483),BA1,Stand Alone,"The minor allele frequency using the filtering allele frequency of either exomes or genomes of continental populations in gnomAD is
128
+ ≥0.00697
129
+
130
+ for AR variants
131
+ . All populations used should have at least 2000 alleles and >1 observation. 
132
+
133
+
134
+ BA1 exclusion variants
135
+ (well-known pathogenic variants that are above the specified BA1 or BS1 threshold) are as follows: 
136
+
137
+
138
+ NM_000540.3:c.6721C>T (p.Arg2241Ter)
139
+
140
+
141
+ NM_000540.3:c.10348-6C>G","Disease-specific,Gene-specific"
142
+ RYR1 (HGNC:10483),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
143
+ RYR1 (HGNC:10483),BS1,Strong,"The minor allele frequency using the filtering allele frequency of either exomes or genomes of continental populations in gnomAD is 
144
+ ≥0.0000006 for AR
145
+
146
+ variants
147
+ . All populations used should have at least 2000 alleles and >1 observation. 
148
+
149
+
150
+ BA1/BS1 exclusion variants
151
+ (well-known pathogenic variants that are above the specified BA1 or BS1 threshold) are as follows: 
152
+
153
+
154
+ NM_000540.3:c.6721C>T (p.Arg2241Ter)
155
+
156
+
157
+ NM_000540.3:c.10348-6C>G","Disease-specific,Gene-specific"
158
+ RYR1 (HGNC:10483),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
159
+ RYR1 (HGNC:10483),BS2,Strong,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",No change
160
+ RYR1 (HGNC:10483),BS2,Moderate,No change - use as originally described,No change
161
+ RYR1 (HGNC:10483),BS2,Supporting,No change - use as originally described,No change
162
+ RYR1 (HGNC:10483),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,NA
163
+ RYR1 (HGNC:10483),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
164
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
165
+ RYR1 (HGNC:10483),BS4,Strong,"Lack of segregation in affected members of a family.
166
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",No change
167
+ RYR1 (HGNC:10483),BS4,Moderate,No change - use as originally described,No change
168
+ RYR1 (HGNC:10483),BS4,Supporting,No change - use as originally described,No change
169
+ RYR1 (HGNC:10483),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
170
+ RYR1 (HGNC:10483),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
171
+ RYR1 (HGNC:10483),BP2,Strong,No change - use as originally described,No change
172
+ RYR1 (HGNC:10483),BP2,Moderate,No change - use as originally described,No change
173
+ RYR1 (HGNC:10483),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,No change
174
+ RYR1 (HGNC:10483),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
175
+ RYR1 (HGNC:10483),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
176
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
177
+ RYR1 (HGNC:10483),BP4,Supporting,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc) Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.
178
+
179
+
180
+ BP4 is met if the REVEL score ≤ 0.15 or if the variant is not predicted to impact splicing using SpliceAI.",General recommendation
181
+ RYR1 (HGNC:10483),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
182
+ RYR1 (HGNC:10483),BP5,Strong,No change - use as originally described,No change
183
+ RYR1 (HGNC:10483),BP5,Moderate,No change - use as originally described,No change
184
+ RYR1 (HGNC:10483),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease.,No change
185
+ RYR1 (HGNC:10483),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
186
+ RYR1 (HGNC:10483),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
187
+ RYR1 (HGNC:10483),BP7,Supporting,A synonymous variant for which SpliceAI predicts no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenCongenitalMyopathiesExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforRYR1Version2.0.0_version=2.0.0.csv ADDED
@@ -0,0 +1,185 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ RYR1 (HGNC:10483),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ RYR1 (HGNC:10483),PVS1,Very Strong,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
9
+ Caveats:
10
+
11
+
12
+
13
+
14
+ Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
15
+
16
+
17
+ Use caution interpreting LOF variants at the extreme 3’ end of a gene.
18
+
19
+
20
+ Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
21
+
22
+
23
+ Use caution in the presence of multiple transcripts.","Disease-specific,Gene-specific"
24
+ RYR1 (HGNC:10483),PVS1,Strong,"In-frame deletions or in frame exon-skipping variants in the pore/transmembrane region of RYR1 should be scored at PVS1_strong. (Amino acids 4800-4950, Exons 100-103)",Gene-specific
25
+ RYR1 (HGNC:10483),PVS1,Moderate,See PVS1 flowchart,"Disease-specific,Gene-specific"
26
+ RYR1 (HGNC:10483),PVS1,Supporting,See PVS1 flowchart,"Disease-specific,Gene-specific"
27
+ RYR1 (HGNC:10483),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
28
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
29
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
30
+ RYR1 (HGNC:10483),PS1,Strong,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
31
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
32
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
33
+ RYR1 (HGNC:10483),PS1,Moderate,No change - use as originally described,No change
34
+ RYR1 (HGNC:10483),PS1,Supporting,No change - use as originally described,No change
35
+ RYR1 (HGNC:10483),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
36
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
37
+ RYR1 (HGNC:10483),PS2,Very Strong,No change - use as originally described,No change
38
+ RYR1 (HGNC:10483),PS2,Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
39
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",No change
40
+ RYR1 (HGNC:10483),PS2,Moderate,No change - use as originally described,No change
41
+ RYR1 (HGNC:10483),PS2,Supporting,No change - use as originally described,No change
42
+ RYR1 (HGNC:10483),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
43
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
44
+ RYR1 (HGNC:10483),PS3,Strong,"Strong may only be considered for variant-specific mouse models. Currently, no other assays are applicable at this strength.",Disease-specific
45
+ RYR1 (HGNC:10483),PS3,Supporting,"There are no specified functional assay for AR RYR1-related myopathy. However, it is acceptable to use PS3_Supporting for other functional analyses if
46
+
47
+
48
+
49
+
50
+ The assay has been validated by a known pathogenic and benign variant AND
51
+
52
+
53
+ There is plausible reason that the function the assay is testing relates to the phenotype AND
54
+
55
+
56
+ The assay conditions are likely to mimic the physiological environment.",Gene-specific
57
+ RYR1 (HGNC:10483),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
58
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
59
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",NA
60
+ RYR1 (HGNC:10483),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
61
+ RYR1 (HGNC:10483),PM1,Moderate,The pore/transmembrane region of RYR1 is critical for protein function and missense variants that fall within this region (amino acids 4800-4950),"Disease-specific,Gene-specific"
62
+ RYR1 (HGNC:10483),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
63
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
64
+ RYR1 (HGNC:10483),PM2,Supporting,"PM2_Supporting may be applied if the minor allele frequency in population databases of at least 2000 alleles is 
65
+
66
+
67
+  ≤ 0.00000697 for autosomal recessive","Disease-specific,Gene-specific"
68
+ RYR1 (HGNC:10483),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
69
+ Note: This requires testing of parents (or offspring) to determine phase.",
70
+ RYR1 (HGNC:10483),PM3,Very Strong,4.0 points per the PM3 chart,"Disease-specific,Gene-specific"
71
+ RYR1 (HGNC:10483),PM3,Strong,2.0 points per the PM3 chart,"Disease-specific,Gene-specific"
72
+ RYR1 (HGNC:10483),PM3,Moderate,"For recessive disorders, detected in trans with a pathogenic variant Note: This requires testing of parents (or offspring) to determine phase.
73
+
74
+
75
+ 1.0 points per the PM3 chart","Disease-specific,Gene-specific"
76
+ RYR1 (HGNC:10483),PM3,Supporting,0.5 points per the PM3 chart,"Disease-specific,Gene-specific"
77
+ RYR1 (HGNC:10483),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
78
+ RYR1 (HGNC:10483),PM4,Strong,No change - use as originally described,No change
79
+ RYR1 (HGNC:10483),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
80
+ RYR1 (HGNC:10483),PM4,Supporting,No change - use as originally described,No change
81
+ RYR1 (HGNC:10483),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
82
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
83
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
84
+ RYR1 (HGNC:10483),PM5,Strong,No change - use as originally described,No change
85
+ RYR1 (HGNC:10483),PM5,Moderate,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
86
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
87
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",No change
88
+ RYR1 (HGNC:10483),PM5,Supporting,No change - use as originally described,No change
89
+ RYR1 (HGNC:10483),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
90
+ RYR1 (HGNC:10483),PM6,Strong,No change - use as originally described,No change
91
+ RYR1 (HGNC:10483),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity.",No change
92
+ RYR1 (HGNC:10483),PM6,Supporting,No change - use as originally described,No change
93
+ RYR1 (HGNC:10483),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
94
+ Note: May be used as stronger evidence with increasing segregation data.",
95
+ RYR1 (HGNC:10483),PP1,Strong,See segregation chart,General recommendation
96
+ RYR1 (HGNC:10483),PP1,Moderate,See segregation chart,General recommendation
97
+ RYR1 (HGNC:10483),PP1,Supporting,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease. Note: May be used as stronger evidence with increasing segregation data.
98
+
99
+
100
+ See segregation chart",General recommendation
101
+ RYR1 (HGNC:10483),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
102
+ RYR1 (HGNC:10483),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
103
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
104
+ RYR1 (HGNC:10483),PP3,Supporting,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.). Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.
105
+
106
+
107
+ PP3 is met if the REVEL score ≥ 0.7 or if the variant is predicted to impact splicing using SpliceAI score ≥0.5",General recommendation
108
+ RYR1 (HGNC:10483),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
109
+ RYR1 (HGNC:10483),PP4,Supporting,"To meet PP4, a congenital myopathy testing panel should be performed without identification of other causative variants and AT LEAST TWO of these features should be present
110
+
111
+
112
+
113
+
114
+ Presence on Muscle Biopsy of:
115
+ mini cores or central cores (histology or electron microscopy)
116
+
117
+
118
+ Exercise, heat, or anesthetic induced rhabdomyolysis
119
+
120
+
121
+ Ophthalmoplegia
122
+
123
+
124
+ Characteristic muscle imaging (see Figure 8, Saade et al 2019 PMID: 31060725)","Disease-specific,Gene-specific"
125
+ RYR1 (HGNC:10483),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
126
+ RYR1 (HGNC:10483),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
127
+ RYR1 (HGNC:10483),BA1,Stand Alone,"The minor allele frequency using the filtering allele frequency of either exomes or genomes of continental populations in gnomAD is
128
+ ≥0.00697
129
+
130
+ for AR variants
131
+ . All populations used should have at least 2000 alleles and >1 observation. 
132
+
133
+
134
+ BA1 exclusion variants
135
+ (well-known pathogenic variants that are above the specified BA1 or BS1 threshold) are as follows: 
136
+
137
+
138
+ NM_000540.3:c.6721C>T (p.Arg2241Ter)
139
+
140
+
141
+ NM_000540.3:c.10348-6C>G","Disease-specific,Gene-specific"
142
+ RYR1 (HGNC:10483),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
143
+ RYR1 (HGNC:10483),BS1,Strong,"The minor allele frequency using the filtering allele frequency of either exomes or genomes of continental populations in gnomAD is 
144
+ ≥0.000697 for AR variants
145
+ . All populations used should have at least 2000 alleles and >1 observation. 
146
+
147
+
148
+ BA1/BS1 exclusion variants
149
+ (well-known pathogenic variants that are above the specified BA1 or BS1 threshold) are as follows: 
150
+
151
+
152
+ NM_000540.3:c.6721C>T (p.Arg2241Ter)
153
+
154
+
155
+ NM_000540.3:c.10348-6C>G","Disease-specific,Gene-specific"
156
+ RYR1 (HGNC:10483),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
157
+ RYR1 (HGNC:10483),BS2,Strong,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",No change
158
+ RYR1 (HGNC:10483),BS2,Moderate,No change - use as originally described,No change
159
+ RYR1 (HGNC:10483),BS2,Supporting,No change - use as originally described,No change
160
+ RYR1 (HGNC:10483),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,NA
161
+ RYR1 (HGNC:10483),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
162
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
163
+ RYR1 (HGNC:10483),BS4,Strong,"Lack of segregation in affected members of a family.
164
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",No change
165
+ RYR1 (HGNC:10483),BS4,Moderate,No change - use as originally described,No change
166
+ RYR1 (HGNC:10483),BS4,Supporting,No change - use as originally described,No change
167
+ RYR1 (HGNC:10483),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
168
+ RYR1 (HGNC:10483),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
169
+ RYR1 (HGNC:10483),BP2,Strong,No change - use as originally described,No change
170
+ RYR1 (HGNC:10483),BP2,Moderate,No change - use as originally described,No change
171
+ RYR1 (HGNC:10483),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,No change
172
+ RYR1 (HGNC:10483),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
173
+ RYR1 (HGNC:10483),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
174
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
175
+ RYR1 (HGNC:10483),BP4,Supporting,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc) Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.
176
+
177
+
178
+ BP4 is met if the REVEL score ≤ 0.15 or if the variant is not predicted to impact splicing using SpliceAI.",General recommendation
179
+ RYR1 (HGNC:10483),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
180
+ RYR1 (HGNC:10483),BP5,Strong,No change - use as originally described,No change
181
+ RYR1 (HGNC:10483),BP5,Moderate,No change - use as originally described,No change
182
+ RYR1 (HGNC:10483),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease.,No change
183
+ RYR1 (HGNC:10483),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
184
+ RYR1 (HGNC:10483),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
185
+ RYR1 (HGNC:10483),BP7,Supporting,A synonymous variant for which SpliceAI predicts no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1.1.0_version=1.1.0.csv ADDED
@@ -0,0 +1,205 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ DICER1 (HGNC:17098),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ DICER1 (HGNC:17098),PVS1,Very Strong,"Follow SVI guidance, using DICER1-specific information.
9
+ Per the PVS1 workflow guidance provided in Tayoun et al. 2018 (PMID 30192042), the following will apply:
10
+
11
+
12
+
13
+
14
+ Nonsense or frameshift variants:
15
+
16
+
17
+ PVS1 applies to variants predicted to result in nonsense-mediated decay (NMD); the predicted NMD cutoff for DICER1 occurs at p.Pro1850.
18
+
19
+
20
+ PVS1_Moderate applies to variants resulting in protein truncation 3’ of this cutoff
21
+
22
+
23
+ Canonical splice variants (+/- 1,2 intronic positions): PVS1 applies with the following exceptions:
24
+
25
+
26
+ Exon 10 SDS/SAS: PVS1_Strong (in-frame but exon includes >10% protein)
27
+
28
+
29
+ Exons 5, 15, 18, 22 SDS/SAS: PVS1_Moderate (in-frame and each <10% of protein)
30
+
31
+
32
+ Exon 27 SAS: PVS1_Moderate (final exon)
33
+
34
+
35
+ Exon 1: no criteria (non-coding)
36
+
37
+
38
+ Variants that disrupt the translation start site (p.M1?): no criteria applied given p.M1 is not highly conserved, there are three in-frame possible alternate start codons (p.Met11, p.Met17, p.Met24), and multiple lab cases of p.Met1? without
39
+ DICER1 phenotype.
40
+ SDS = splice donor site; SAS = splice acceptor site.
41
+ Refer to PS3 weight guidelines when a variant meets criterion for application of both PVS1 and PS3.
42
+ A disease-specific PVS1 decision tree incorporating the above bullets is also included at the end of this document as an additional curation tool.","Disease-specific,General recommendation"
43
+ DICER1 (HGNC:17098),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
44
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
45
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
46
+ DICER1 (HGNC:17098),PS1,Strong,"For same AA change, must confirm there is no difference in splicing using RNA data or in-silico modeling data (concordance of MaxEntScan and SpliceAI). For non-canonical intronic splicing variants at same nucleotide should have equal or worse splicing impact.
47
+ This rule code can only be used to compare variants asserted as pathogenic by the ClinGen DICER1 VCEP. Likely pathogenic changes do not apply.",General recommendation
48
+ DICER1 (HGNC:17098),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
49
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
50
+ DICER1 (HGNC:17098),PS2,Very Strong,"≥4 de novo points.
51
+ De novo points should be tallied using the simplified table for tallying
52
+ proband points and used to determine the applied strength of PS2, consistent
53
+ with SVI guidance. To avoid redundancy and increase consistency, the EP has
54
+ opted to drop PM6 and exclusively use PS2 for de novo evidence.",Strength
55
+ DICER1 (HGNC:17098),PS2,Strong,"≥2 but less than 4 de novo points.
56
+ De novo points should be tallied using the simplified table for tallying
57
+ proband points and used to determine the applied strength of PS2, consistent
58
+ with SVI guidance. To avoid redundancy and increase consistency, the EP has
59
+ opted to drop PM6 and exclusively use PS2 for de novo evidence.",Strength
60
+ DICER1 (HGNC:17098),PS2,Moderate,"≥1 but less than 2 de novo points.
61
+ De novo points should be tallied using the simplified table for tallying proband points and used to determine the applied strength of PS2, consistent with SVI guidance. To avoid redundancy and increase consistency, the EP has opted to drop PM6 and exclusively use PS2 for de novo evidence.",General recommendation
62
+ DICER1 (HGNC:17098),PS2,Supporting,"≥0.5 but less than 1 de novo points.
63
+ De novo points should be tallied using the simplified table for tallying
64
+ proband points and used to determine the applied strength of PS2, consistent with SVI guidance. To avoid redundancy and increase consistency, the EP has opted to drop PM6 and exclusively use PS2 for de novo evidence.",Strength
65
+ DICER1 (HGNC:17098),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
66
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
67
+ DICER1 (HGNC:17098),PS3,Strong,"RNA assay shows splicing impact that is out-of-frame, in-frame ≥193 residues, or in-frame with RNase IIIb disruption.
68
+ (PS3_Moderate if PVS1_Strong is applied).
69
+ This rule should be used and weighted appropriately for variants with functional evidence of a splicing impact and/or reduced DICER1 ability to cleave pre-miRNA. Follow SVI guidance regarding control numbers for functional studies. Do not apply PS3 at any strength if PVS1 is applied at full strength.",Disease-specific
70
+ DICER1 (HGNC:17098),PS3,Moderate,"RNA assay shows in-frame splicing impact with change in protein length <193 residues AND RNase IIIb domain not disrupted.
71
+ This rule should be used and weighted appropriately for variants with functional evidence of a splicing impact and/or reduced DICER1 ability to cleave pre-miRNA. Follow SVI guidance regarding control numbers for functional studies. Do not apply PS3 at any strength if PVS1 is applied at full strength.","Disease-specific,General recommendation"
72
+ DICER1 (HGNC:17098),PS3,Supporting,"In vitro cleavage assay shows failure or severely reduced capacity to produce either 5p or 3p microRNAs from a premiRNA (positive and negative controls also performed).
73
+ This rule should be used and weighted appropriately for variants with functional evidence of a splicing impact and/or reduced DICER1 ability to cleave pre-miRNA. Follow SVI guidance regarding control numbers for functional studies. Do not apply PS3 at any strength if PVS1 is applied at full strength.","Disease-specific,Strength"
74
+ DICER1 (HGNC:17098),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
75
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
76
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
77
+ DICER1 (HGNC:17098),PS4,Strong,"≥4 phenotype points.
78
+ Unrelated probands may contribute up to 1 point each based on phenotype (see Tables 2 & 3 in ruleset)
79
+ Caveats:
80
+
81
+
82
+
83
+
84
+ Do not apply PS4 if variant meets BA1/BS1 criteria.
85
+
86
+
87
+ Do not apply points for a phenotype in an individual with a likely pathogenic germline variant in a second gene that could have reasonably contributed to the phenotype (e.g. Wilms tumor in an individual with a P/LP WT1 variant).
88
+
89
+
90
+ Do not apply points for a proband whose tumor sequencing is consistent with a likely sporadic event (i.e. sequencing reveals a somatic, VCEPcurated, non-hotspot, likely pathogenic DICER1 variant in addition to a somatic hotspot variant and the germline variant under assessment). Of note, DICER1 tumors that consistently or occasionally follow a classical 2- hit hypothesis (i.e. LOF of both alleles) are exempt from this caveat. For example, identification of a somatic pathogenic non-hotspot DICER1 variant in pineoblastoma (PMID: 25022261), pituitary blastoma (PMID: 24839956), and lung cysts or cystic nephroma lacking mesenchymal cells (PMIDs: 25500911, 25978641) should not exclude the proband from PS4.",General recommendation
91
+ DICER1 (HGNC:17098),PS4,Moderate,"2 – 3.5 phenotype points.
92
+ Unrelated probands may contribute up to 1 point each based on phenotype (see Tables 2 & 3 in ruleset)
93
+ Caveats:
94
+
95
+
96
+
97
+
98
+ Do not apply PS4 if variant meets BA1/BS1 criteria.
99
+
100
+
101
+ Do not apply points for a phenotype in an individual with a likely pathogenic germline variant in a second gene that could have reasonably contributed to the phenotype (e.g. Wilms tumor in an individual with a P/LP WT1 variant).
102
+
103
+
104
+ Do not apply points for a proband whose tumor sequencing is consistent with a likely sporadic event (i.e. sequencing reveals a somatic, VCEPcurated, non-hotspot, likely pathogenic DICER1 variant in addition to a somatic hotspot variant and the germline variant under assessment). Of note, DICER1 tumors that consistently or occasionally follow a classical 2- hit hypothesis (i.e. LOF of both alleles) are exempt from this caveat. For example, identification of a somatic pathogenic non-hotspot DICER1 variant in pineoblastoma (PMID: 25022261), pituitary blastoma (PMID: 24839956), and lung cysts or cystic nephroma lacking mesenchymal cells (PMIDs: 25500911, 25978641) should not exclude the proband from PS4.",Strength
105
+ DICER1 (HGNC:17098),PS4,Supporting,"1 – 1.5 phenotype points.
106
+ Unrelated probands may contribute up to 1 point each based on phenotype (see Tables 2 & 3 in ruleset)
107
+ Caveats:
108
+
109
+
110
+
111
+
112
+ Do not apply PS4 if variant meets BA1/BS1 criteria.
113
+
114
+
115
+ Do not apply points for a phenotype in an individual with a likely pathogenic germline variant in a second gene that could have reasonably contributed to the phenotype (e.g. Wilms tumor in an individual with a P/LP WT1 variant).
116
+
117
+
118
+ Do not apply points for a proband whose tumor sequencing is consistent with a likely sporadic event (i.e. sequencing reveals a somatic, VCEPcurated, non-hotspot, likely pathogenic DICER1 variant in addition to a somatic hotspot variant and the germline variant under assessment). Of note, DICER1 tumors that consistently or occasionally follow a classical 2- hit hypothesis (i.e. LOF of both alleles) are exempt from this caveat. For example, identification of a somatic pathogenic non-hotspot DICER1 variant in pineoblastoma (PMID: 25022261), pituitary blastoma (PMID: 24839956), and lung cysts or cystic nephroma lacking mesenchymal cells (PMIDs: 25500911, 25978641) should not exclude the proband from PS4.",Strength
119
+ DICER1 (HGNC:17098),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
120
+ DICER1 (HGNC:17098),PM1,Moderate,"Putative missense variants at residues affecting metal ion-binding: codons p.S1344, p.E1705, p.D1709, p.D1713, p.G1809, p.D1810, p.E1813",Disease-specific
121
+ DICER1 (HGNC:17098),PM1,Supporting,"Putative missense variants at residues in the RNase IIIb domain (p.Y1682 – p.S1846), besides the metal ion-binding residues (see PM1).",Strength
122
+ DICER1 (HGNC:17098),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
123
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
124
+ DICER1 (HGNC:17098),PM2,Supporting,Allele frequency <0.000005 across gnomAD (non-cancer) with no more than one allele in any subpopulation and at least 20x coverage.,"Disease-specific,Strength"
125
+ DICER1 (HGNC:17098),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
126
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
127
+ DICER1 (HGNC:17098),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
128
+ DICER1 (HGNC:17098),PM4,Moderate,In-frame indels with a residue within the RNase IIIb domain (p.Y1682 – p.S1846).,Disease-specific
129
+ DICER1 (HGNC:17098),PM4,Supporting,In-frame indels outside of the RNase IIIb domain (p.Y1682 – p.S1846) and repeat regions (p.D606-p.D609; p.E1418-p.E1420; p.E1422-p.E1425).,Strength
130
+ DICER1 (HGNC:17098),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
131
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
132
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
133
+ DICER1 (HGNC:17098),PM5,Moderate,Missense variant under evaluation should have equal or worse Grantham score. Splicing should be ruled out with either RNA data or agreement in splicing predictors (MaxEntScan and SpliceAI) that show no splicing effects. The other variant must be interpreted as pathogenic by the ClinGen DICER1 VCEP. Likely pathogenic changes do not apply. This rule cannot be applied in combination with PM1 or PS1.,General recommendation
134
+ DICER1 (HGNC:17098),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
135
+ DICER1 (HGNC:17098),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
136
+ Note: May be used as stronger evidence with increasing segregation data.",
137
+ DICER1 (HGNC:17098),PP1,Strong,"≥7 meioses across ≥2 families.
138
+ Phenotype-positive individuals should have high, moderate, or low-specificity phenotypes (see phenotype table). (Caveat: segregation with a single low-specificity phenotype across multiple individuals (e.g. familial Wilms tumor) does not fulfill PP1.)
139
+ Do not apply PP1 if variant meets BA1/BS1 criteria.",Strength
140
+ DICER1 (HGNC:17098),PP1,Moderate,"5 – 6 meioses across ≥1 family.
141
+ Phenotype-positive individuals should have high, moderate, or low-specificity phenotypes (see phenotype table). (Caveat: segregation with a single low-specificity
142
+ phenotype across multiple individuals (e.g. familial Wilms tumor) does not fulfill PP1.)
143
+ Do not apply PP1 if variant meets BA1/BS1 criteria.",Strength
144
+ DICER1 (HGNC:17098),PP1,Supporting,"3 – 4 meioses across ≥1 family.
145
+ Phenotype-positive individuals should have high, moderate, or low-specificity phenotypes (see phenotype table). (Caveat: segregation with a single low-specificity phenotype across multiple individuals (e.g. familial Wilms tumor) does not fulfill PP1.)
146
+ Do not apply PP1 if variant meets BA1/BS1 criteria.",General recommendation
147
+ DICER1 (HGNC:17098),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
148
+ DICER1 (HGNC:17098),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
149
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
150
+ DICER1 (HGNC:17098),PP3,Supporting,"For missense variants, REVEL score ≥ 0.75 OR agreement in splicing predictors predict splicing effects. For splicing variants, concordance of MaxEntScan and SpliceAI.",Disease-specific
151
+ DICER1 (HGNC:17098),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
152
+ DICER1 (HGNC:17098),PP4,Supporting,"Somatic tumor testing identifies somatic hotspot second hit and no additional somatic LOF variants.
153
+ Tumor testing (PMID: 30311369) of a neoplasm with known DICER1 association in a proband who carries the germline variant under evaluation reveals the
154
+ following:
155
+
156
+
157
+
158
+
159
+ A previously reported somatic second hit of DICER1 in an RNase IIIb-disrupting “hotspot” codon (p.S1344, p.E1705, p.D1709, p.D1713, p.G1809, p.D1810, or
160
+ p.E1813) AND
161
+
162
+
163
+ Retention of the germline DICER1 variant under evaluation.
164
+ PP4 is NOT applicable if:
165
+
166
+
167
+ The germline variant is a missense variant in one of the seven RNase IIIb “hotspot” codons (see PM1), OR
168
+
169
+
170
+ Somatic sequencing reveals additional DICER1 non-hotspot variants (could be consistent with sporadic tumorigenesis).",Disease-specific
171
+ DICER1 (HGNC:17098),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
172
+ DICER1 (HGNC:17098),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
173
+ DICER1 (HGNC:17098),BA1,Stand Alone,"Frequency >0.003 (0.3%) in gnomAD (non-cancer) subpopulations. Subpopulations must have >2,000 alleles tested and a minimum of 5 alleles present.",Disease-specific
174
+ DICER1 (HGNC:17098),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
175
+ DICER1 (HGNC:17098),BS1,Strong,"Frequency >0.0003 (0.03%) in gnomAD (non-cancer) subpopulations. Subpopulations must have >2,000 alleles tested and a minimum of 5 alleles present.",Disease-specific
176
+ DICER1 (HGNC:17098),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
177
+ DICER1 (HGNC:17098),BS2,Strong,"40+ unrelated females from a single source are tumor-free through age 50 (caveat: ratio of BS2-eligible females to PS4-eligible probands must be ≥ 40:1)
178
+ OR 2+ observations of homozygosity in healthy individuals
179
+ OR 1+ observation(s) of homozygosity in a healthy individual with status confirmed by parental testing.",Disease-specific
180
+ DICER1 (HGNC:17098),BS2,Supporting,"10+ unrelated females from a single source are tumor-free through age 50 (caveat: ratio of BS2-eligible females to PS4-eligible probands must be ≥ 10:1)
181
+ OR 2+ observations of homozygosity in individuals lacking clinical information",Disease-specific
182
+ DICER1 (HGNC:17098),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
183
+ DICER1 (HGNC:17098),BS3,Strong,"This rule should be used and weighted appropriately for variants with functional evidence of no splicing impact and/or no reduced DICER1 ability to cleave pre-miRNA. Follow SVI guidance regarding control numbers for functional studies.
184
+ For intronic or synonymous variants, no splicing impact observed via RNA assay. (Should be observed more than once.)",Disease-specific
185
+ DICER1 (HGNC:17098),BS3,Supporting,"This rule should be used and weighted appropriately for variants with functional evidence of no splicing impact and/or no reduced DICER1 ability to cleave pre-miRNA. Follow SVI guidance regarding control numbers for functional studies.
186
+ An in vitro cleavage assay must demonstrate the variant produces both 5p and 3p microRNAs from a pre-miRNA (positive and negative controls also performed). An example of an appropriate assay to which criteria could be applied is Wu et al. 2018 (PMID: 28862265).",Disease-specific
187
+ DICER1 (HGNC:17098),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
188
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
189
+ DICER1 (HGNC:17098),BS4,Strong,"Family members should be phenotype-positive (must be high- or moderatespecificity phenotype; see phenotype table), genotype-negative 1st, 2nd, or 3rd degree relatives of the proband.",General recommendation
190
+ DICER1 (HGNC:17098),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
191
+ DICER1 (HGNC:17098),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
192
+ DICER1 (HGNC:17098),BP2,Supporting,"≥1 observation in trans with P/LP DICER1 variant or ≥3
193
+ observations in cis or phase unknown with 2+ different
194
+ P/LP DICER1 variants.",Disease-specific
195
+ DICER1 (HGNC:17098),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
196
+ DICER1 (HGNC:17098),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
197
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
198
+ DICER1 (HGNC:17098),BP4,Supporting,"For missense variants, REVEL score < 0.50 and agreement in splicing predictors that no splicing effects are predicted. For synonymous/intronic/non-coding variants concordance of MaxEntScan and SpliceAI.",Disease-specific
199
+ DICER1 (HGNC:17098),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
200
+ DICER1 (HGNC:17098),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
201
+ DICER1 (HGNC:17098),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
202
+ DICER1 (HGNC:17098),BP7,Supporting,"Silent variant
203
+ OR Intronic variant at or beyond +7 to -21 positions
204
+ OR Other intronic or non-coding variant if the variant is the reference nucleotide in ≥1 primate and/or ≥4 mammalian species.
205
+ Caveat: Variant must meet BP4 to apply BP7",General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1.2.0_version=1.2.0.csv ADDED
@@ -0,0 +1,139 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ DICER1 (HGNC:17098),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ DICER1 (HGNC:17098),PVS1,Very Strong,"Follow SVI guidance, using DICER1-specific information. Per the PVS1 workflow guidance provided in Tayoun et al. 2018
9
+ 1
10
+ , the following will apply:
11
+
12
+
13
+
14
+
15
+ Nonsense or frameshift variants:
16
+
17
+
18
+ PVS1 applies to variants predicted to result in nonsense-mediated decay (NMD); the predicted NMD cutoff for DICER1 occurs at p.Pro1850.
19
+
20
+
21
+ PVS1_Moderate applies to variants resulting in protein truncation 3’ of this cutoff
22
+
23
+
24
+ Canonical splice variants (+/- 1,2 intronic positions): PVS1 applies with the following exceptions:
25
+
26
+
27
+ Exon 10 SDS/SAS: PVS1_Strong (in-frame but exon includes >10% protein)
28
+
29
+
30
+ Exons 5, 15, 18, 22 SDS/SAS: PVS1_Moderate (in-frame and each <10% of protein)
31
+
32
+
33
+ Exon 27 SAS: PVS1_Moderate (final exon)
34
+
35
+
36
+ Exon 1: no criteria (non-coding)
37
+
38
+
39
+ Variants that disrupt the translation start site (p.M1?): no criteria applied given p.M1 is not highly conserved, there are three in-frame possible alternate start codons (p.Met11, p.Met17, p.Met24), and multiple lab cases of p.Met1? without DICER1 phenotype. SDS = splice donor site; SAS = splice acceptor site. Refer to PS3 weight guidelines when a variant meets criterion for application of both PVS1 and PS3. A disease-specific PVS1 decision tree incorporating the above bullets is also included at the end of this document as an additional curation tool.","Disease-specific,General recommendation"
40
+ DICER1 (HGNC:17098),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
41
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
42
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
43
+ DICER1 (HGNC:17098),PS1,Strong,"For same AA change, must confirm there is no difference in splicing using RNA data or in-silico modeling data (concordance of MaxEntScan and SpliceAI). For non-canonical intronic splicing variants at same nucleotide should have equal or worse splicing impact.
44
+ This rule code can only be used to compare variants asserted as pathogenic by the ClinGen DICER1 VCEP. Likely pathogenic changes do not apply.",General recommendation
45
+ DICER1 (HGNC:17098),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
46
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
47
+ DICER1 (HGNC:17098),PS2,Very Strong,≥4 de novo points,Strength
48
+ DICER1 (HGNC:17098),PS2,Strong,≥2 but less than 4 de novo points,Strength
49
+ DICER1 (HGNC:17098),PS2,Moderate,≥1 but less than 2 de novo points,General recommendation
50
+ DICER1 (HGNC:17098),PS2,Supporting,≥0.5 but less than 1 de novo points,Strength
51
+ DICER1 (HGNC:17098),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
52
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
53
+ DICER1 (HGNC:17098),PS3,Strong,"RNA assay shows splicing impact that is out-of-frame, in-frame ≥193 residues, or in-frame with RNase IIIb disruption. (PS3_Moderate if PVS1_Strong is applied).",Disease-specific
54
+ DICER1 (HGNC:17098),PS3,Moderate,RNA assay shows in-frame splicing impact with change in protein length <193 residues AND RNase IIIb domain not disrupted.,"Disease-specific,General recommendation"
55
+ DICER1 (HGNC:17098),PS3,Supporting,In vitro cleavage assay shows failure or severely reduced capacity to produce either 5p or 3p microRNAs from a premiRNA (positive and negative controls also performed).,"Disease-specific,Strength"
56
+ DICER1 (HGNC:17098),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
57
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
58
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
59
+ DICER1 (HGNC:17098),PS4,Strong,≥4 phenotype points,General recommendation
60
+ DICER1 (HGNC:17098),PS4,Moderate,2 – 3.5 phenotype points,Strength
61
+ DICER1 (HGNC:17098),PS4,Supporting,1 – 1.5 phenotype points,Strength
62
+ DICER1 (HGNC:17098),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
63
+ DICER1 (HGNC:17098),PM1,Moderate,"Putative missense variants at residues affecting metal ion-binding: codons p.S1344, p.E1705, p.D1709, p.D1713, p.G1809, p.D1810, p.E1813",Disease-specific
64
+ DICER1 (HGNC:17098),PM1,Supporting,"Putative missense variants at residues in the RNase IIIb domain (p.Y1682 – p.S1846), besides the metal ion-binding residues (see PM1).",Strength
65
+ DICER1 (HGNC:17098),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
66
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
67
+ DICER1 (HGNC:17098),PM2,Supporting,Allele frequency <0.000005 across gnomAD (non-cancer) with no more than one allele in any subpopulation and at least 20x coverage.,"Disease-specific,Strength"
68
+ DICER1 (HGNC:17098),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
69
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
70
+ DICER1 (HGNC:17098),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
71
+ DICER1 (HGNC:17098),PM4,Moderate,In-frame indels with a residue within the RNase IIIb domain (p.Y1682 – p.S1846).,Disease-specific
72
+ DICER1 (HGNC:17098),PM4,Supporting,In-frame indels outside of the RNase IIIb domain (p.Y1682 – p.S1846) and repeat regions (p.D606-p.D609; p.E1418-p.E1420; p.E1422-p.E1425).,Strength
73
+ DICER1 (HGNC:17098),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
74
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
75
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
76
+ DICER1 (HGNC:17098),PM5,Moderate,Missense variant under evaluation should have equal or worse Grantham score. Splicing should be ruled out with either RNA data or agreement in splicing predictors (MaxEntScan and SpliceAI) that show no splicing effects. The other variant must be interpreted as pathogenic by the ClinGen DICER1 VCEP. Likely pathogenic changes do not apply. This rule cannot be applied in combination with PM1 or PS1.,General recommendation
77
+ DICER1 (HGNC:17098),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
78
+ DICER1 (HGNC:17098),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
79
+ Note: May be used as stronger evidence with increasing segregation data.",
80
+ DICER1 (HGNC:17098),PP1,Strong,≥7 meioses across ≥2 families,Strength
81
+ DICER1 (HGNC:17098),PP1,Moderate,5 – 6 meioses across ≥1 family,Strength
82
+ DICER1 (HGNC:17098),PP1,Supporting,3 – 4 meioses across ≥1 family,General recommendation
83
+ DICER1 (HGNC:17098),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
84
+ DICER1 (HGNC:17098),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
85
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
86
+ DICER1 (HGNC:17098),PP3,Supporting,"For missense variants, REVEL score ≥ 0.75 OR agreement in splicing predictors predict splicing effects. For splicing variants, concordance of MaxEntScan and SpliceAI.",Disease-specific
87
+ DICER1 (HGNC:17098),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
88
+ DICER1 (HGNC:17098),PP4,Supporting,"Somatic tumor testing identifies somatic hotspot second hit and no additional somatic LOF variants. Tumor testing
89
+ 6
90
+ of a neoplasm with known DICER1 association in a proband who carries the germline variant under evaluation reveals the following:
91
+
92
+
93
+
94
+
95
+ A previously reported somatic second hit of DICER1 in an RNase IIIb-disrupting “hotspot” codon (p.S1344, p.E1705, p.D1709, p.D1713, p.G1809, p.D1810, or p.E1813) AND
96
+
97
+
98
+ Retention of the germline DICER1 variant under evaluation. PP4 is NOT applicable if:
99
+
100
+
101
+ The germline variant is a missense variant in one of the seven RNase IIIb “hotspot” codons (see PM1), OR
102
+
103
+
104
+ Somatic sequencing reveals additional DICER1 non-hotspot variants (could be consistent with sporadic tumorigenesis).",Disease-specific
105
+ DICER1 (HGNC:17098),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
106
+ DICER1 (HGNC:17098),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
107
+ DICER1 (HGNC:17098),BA1,Stand Alone,"Frequency >0.003 (0.3%) in gnomAD (non-cancer) subpopulations. Subpopulations must have >2,000 alleles tested and a minimum of 5 alleles present.",Disease-specific
108
+ DICER1 (HGNC:17098),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
109
+ DICER1 (HGNC:17098),BS1,Strong,"Frequency >0.0003 (0.03%) in gnomAD (non-cancer) subpopulations. Subpopulations must have >2,000 alleles tested and a minimum of 5 alleles present.",Disease-specific
110
+ DICER1 (HGNC:17098),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
111
+ DICER1 (HGNC:17098),BS2,Strong,"40+ unrelated females from a single source are tumor-free through age 50 (caveat: ratio of BS2-eligible females to PS4-eligible probands must be ≥ 40:1)
112
+ OR 2+ observations of homozygosity in healthy individuals
113
+ OR 1+ observation(s) of homozygosity in a healthy individual with status confirmed by parental testing.",Disease-specific
114
+ DICER1 (HGNC:17098),BS2,Supporting,"10+ unrelated females from a single source are tumor-free through age 50 (caveat: ratio of BS2-eligible females to PS4-eligible probands must be ≥ 10:1)
115
+ OR 2+ observations of homozygosity in individuals lacking clinical information",Disease-specific
116
+ DICER1 (HGNC:17098),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
117
+ DICER1 (HGNC:17098),BS3,Strong,"For intronic or synonymous variants, no splicing impact observed via RNA assay. (Should be observed more than once.)",Disease-specific
118
+ DICER1 (HGNC:17098),BS3,Supporting,"An in vitro cleavage assay must demonstrate the variant produces both 5p and 3p microRNAs from a pre-miRNA (positive and negative controls also performed). An example of an appropriate assay to which criteria could be applied is Wu et al. 2018
119
+ 7
120
+ .",Disease-specific
121
+ DICER1 (HGNC:17098),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
122
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
123
+ DICER1 (HGNC:17098),BS4,Strong,"Family members should be phenotype-positive (must be high- or moderatespecificity phenotype; see phenotype table), genotype-negative 1st, 2nd, or 3rd degree relatives of the proband.",General recommendation
124
+ DICER1 (HGNC:17098),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
125
+ DICER1 (HGNC:17098),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
126
+ DICER1 (HGNC:17098),BP2,Supporting,"≥1 observation in trans with P/LP DICER1 variant or ≥3
127
+ observations in cis or phase unknown with 2+ different
128
+ P/LP DICER1 variants.",Disease-specific
129
+ DICER1 (HGNC:17098),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
130
+ DICER1 (HGNC:17098),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
131
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
132
+ DICER1 (HGNC:17098),BP4,Supporting,"For missense variants, REVEL score < 0.50 and agreement in splicing predictors that no splicing effects are predicted. For synonymous/intronic/non-coding variants concordance of MaxEntScan and SpliceAI.",Disease-specific
133
+ DICER1 (HGNC:17098),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
134
+ DICER1 (HGNC:17098),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
135
+ DICER1 (HGNC:17098),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
136
+ DICER1 (HGNC:17098),BP7,Supporting,"Silent variant
137
+ OR Intronic variant at or beyond +7 to -21 positions
138
+ OR Other intronic or non-coding variant if the variant is the reference nucleotide in ≥1 primate and/or ≥4 mammalian species.
139
+ Caveat: Variant must meet BP4 to apply BP7",General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1.3.0_version=1.3.0.csv ADDED
@@ -0,0 +1,139 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ DICER1 (HGNC:17098),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ DICER1 (HGNC:17098),PVS1,Very Strong,"Follow SVI guidance, using DICER1-specific information. Per the PVS1 workflow guidance provided in Tayoun et al. 2018
9
+ 1
10
+ , the following will apply:
11
+
12
+
13
+
14
+
15
+ Nonsense or frameshift variants:
16
+
17
+
18
+ PVS1 applies to variants predicted to result in nonsense-mediated decay (NMD); the predicted NMD cutoff for DICER1 occurs at p.Pro1850.
19
+
20
+
21
+ PVS1_Moderate applies to variants resulting in protein truncation 3’ of this cutoff
22
+
23
+
24
+ Canonical splice variants (+/- 1,2 intronic positions): PVS1 applies with the following exceptions:
25
+
26
+
27
+ Exon 10 SDS/SAS: PVS1_Strong (in-frame but exon includes >10% protein)
28
+
29
+
30
+ Exons 5, 15, 18, 22 SDS/SAS: PVS1_Moderate (in-frame and each <10% of protein)
31
+
32
+
33
+ Exon 27 SAS: PVS1_Moderate (final exon)
34
+
35
+
36
+ Exon 1: no criteria (non-coding)
37
+
38
+
39
+ Variants that disrupt the translation start site (p.M1?): no criteria applied given p.M1 is not highly conserved, there are three in-frame possible alternate start codons (p.Met11, p.Met17, p.Met24), and multiple lab cases of p.Met1? without DICER1 phenotype. SDS = splice donor site; SAS = splice acceptor site. Refer to PS3 weight guidelines when a variant meets criterion for application of both PVS1 and PS3. A disease-specific PVS1 decision tree incorporating the above bullets is also included at the end of this document as an additional curation tool.","Disease-specific,General recommendation"
40
+ DICER1 (HGNC:17098),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
41
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
42
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
43
+ DICER1 (HGNC:17098),PS1,Strong,"For same AA change, must confirm there is no difference in splicing using RNA data or in-silico modeling data (concordance of MaxEntScan and SpliceAI). For non-canonical intronic splicing variants at same nucleotide should have equal or worse splicing impact.
44
+ This rule code can only be used to compare variants asserted as pathogenic by the ClinGen DICER1 VCEP. Likely pathogenic changes do not apply.",General recommendation
45
+ DICER1 (HGNC:17098),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
46
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
47
+ DICER1 (HGNC:17098),PS2,Very Strong,≥4 de novo points,Strength
48
+ DICER1 (HGNC:17098),PS2,Strong,≥2 but less than 4 de novo points,Strength
49
+ DICER1 (HGNC:17098),PS2,Moderate,≥1 but less than 2 de novo points,General recommendation
50
+ DICER1 (HGNC:17098),PS2,Supporting,≥0.5 but less than 1 de novo points,Strength
51
+ DICER1 (HGNC:17098),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
52
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
53
+ DICER1 (HGNC:17098),PS3,Strong,"RNA assay shows splicing impact that is out-of-frame, in-frame ≥193 residues, or in-frame with RNase IIIb disruption. (PS3_Moderate if PVS1_Strong is applied).",Disease-specific
54
+ DICER1 (HGNC:17098),PS3,Moderate,RNA assay shows in-frame splicing impact with change in protein length <193 residues AND RNase IIIb domain not disrupted.,"Disease-specific,General recommendation"
55
+ DICER1 (HGNC:17098),PS3,Supporting,In vitro cleavage assay shows failure or severely reduced capacity to produce either 5p or 3p microRNAs from a premiRNA (positive and negative controls also performed).,"Disease-specific,Strength"
56
+ DICER1 (HGNC:17098),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
57
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
58
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
59
+ DICER1 (HGNC:17098),PS4,Strong,≥4 phenotype points,General recommendation
60
+ DICER1 (HGNC:17098),PS4,Moderate,2 – 3.5 phenotype points,Strength
61
+ DICER1 (HGNC:17098),PS4,Supporting,1 – 1.5 phenotype points,Strength
62
+ DICER1 (HGNC:17098),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
63
+ DICER1 (HGNC:17098),PM1,Moderate,"Putative missense variants at residues affecting metal ion-binding: codons p.S1344, p.E1705, p.D1709, p.D1713, p.G1809, p.D1810, p.E1813",Disease-specific
64
+ DICER1 (HGNC:17098),PM1,Supporting,"Putative missense variants at residues in the RNase IIIb domain (p.Y1682 – p.S1846), besides the metal ion-binding residues (see PM1).",Strength
65
+ DICER1 (HGNC:17098),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
66
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
67
+ DICER1 (HGNC:17098),PM2,Supporting,Allele frequency <0.000005 across gnomAD with no more than one allele in any subpopulation and at least 20x coverage.,"Disease-specific,Strength"
68
+ DICER1 (HGNC:17098),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
69
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
70
+ DICER1 (HGNC:17098),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
71
+ DICER1 (HGNC:17098),PM4,Moderate,In-frame indels with a residue within the RNase IIIb domain (p.Y1682 – p.S1846).,Disease-specific
72
+ DICER1 (HGNC:17098),PM4,Supporting,In-frame indels outside of the RNase IIIb domain (p.Y1682 – p.S1846) and repeat regions (p.D606-p.D609; p.E1418-p.E1420; p.E1422-p.E1425).,Strength
73
+ DICER1 (HGNC:17098),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
74
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
75
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
76
+ DICER1 (HGNC:17098),PM5,Moderate,Missense variant under evaluation should have equal or worse Grantham score. Splicing should be ruled out with either RNA data or agreement in splicing predictors (MaxEntScan and SpliceAI) that show no splicing effects. The other variant must be interpreted as pathogenic by the ClinGen DICER1 VCEP. Likely pathogenic changes do not apply. This rule cannot be applied in combination with PM1 or PS1.,General recommendation
77
+ DICER1 (HGNC:17098),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
78
+ DICER1 (HGNC:17098),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
79
+ Note: May be used as stronger evidence with increasing segregation data.",
80
+ DICER1 (HGNC:17098),PP1,Strong,≥7 meioses across ≥2 families,Strength
81
+ DICER1 (HGNC:17098),PP1,Moderate,5 – 6 meioses across ≥1 family,Strength
82
+ DICER1 (HGNC:17098),PP1,Supporting,3 – 4 meioses across ≥1 family,General recommendation
83
+ DICER1 (HGNC:17098),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
84
+ DICER1 (HGNC:17098),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
85
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
86
+ DICER1 (HGNC:17098),PP3,Supporting,"For missense variants, REVEL score ≥ 0.75 OR agreement in splicing predictors predict splicing effects. For splicing variants, concordance of MaxEntScan and SpliceAI.",Disease-specific
87
+ DICER1 (HGNC:17098),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
88
+ DICER1 (HGNC:17098),PP4,Supporting,"Somatic tumor testing identifies somatic hotspot second hit and no additional somatic LOF variants. Tumor testing
89
+ 6
90
+ of a neoplasm with known DICER1 association in a proband who carries the germline variant under evaluation reveals the following:
91
+
92
+
93
+
94
+
95
+ A previously reported somatic second hit of DICER1 in an RNase IIIb-disrupting “hotspot” codon (p.S1344, p.E1705, p.D1709, p.D1713, p.G1809, p.D1810, or p.E1813) AND
96
+
97
+
98
+ Retention of the germline DICER1 variant under evaluation. PP4 is NOT applicable if:
99
+
100
+
101
+ The germline variant is a missense variant in one of the seven RNase IIIb “hotspot” codons (see PM1), OR
102
+
103
+
104
+ Somatic sequencing reveals additional DICER1 non-hotspot variants (could be consistent with sporadic tumorigenesis).",Disease-specific
105
+ DICER1 (HGNC:17098),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
106
+ DICER1 (HGNC:17098),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
107
+ DICER1 (HGNC:17098),BA1,Stand Alone,"Frequency >0.003 (0.3%) in gnomAD subpopulations. Subpopulations must have >2,000 alleles tested and a minimum of 5 alleles present.",Disease-specific
108
+ DICER1 (HGNC:17098),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
109
+ DICER1 (HGNC:17098),BS1,Strong,"Frequency >0.0003 (0.03%) in gnomAD subpopulations. Subpopulations must have >2,000 alleles tested and a minimum of 5 alleles present.",Disease-specific
110
+ DICER1 (HGNC:17098),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
111
+ DICER1 (HGNC:17098),BS2,Strong,"40+ unrelated females from a single source are tumor-free through age 50 (caveat: ratio of BS2-eligible females to PS4-eligible probands must be ≥ 40:1)
112
+ OR 2+ observations of homozygosity in healthy individuals
113
+ OR 1+ observation(s) of homozygosity in a healthy individual with status confirmed by parental testing.",Disease-specific
114
+ DICER1 (HGNC:17098),BS2,Supporting,"10+ unrelated females from a single source are tumor-free through age 50 (caveat: ratio of BS2-eligible females to PS4-eligible probands must be ≥ 10:1)
115
+ OR 2+ observations of homozygosity in individuals lacking clinical information",Disease-specific
116
+ DICER1 (HGNC:17098),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
117
+ DICER1 (HGNC:17098),BS3,Strong,"For intronic or synonymous variants, no splicing impact observed via RNA assay. (Should be observed more than once.)",Disease-specific
118
+ DICER1 (HGNC:17098),BS3,Supporting,"An in vitro cleavage assay must demonstrate the variant produces both 5p and 3p microRNAs from a pre-miRNA (positive and negative controls also performed). An example of an appropriate assay to which criteria could be applied is Wu et al. 2018
119
+ 7
120
+ .",Disease-specific
121
+ DICER1 (HGNC:17098),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
122
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
123
+ DICER1 (HGNC:17098),BS4,Strong,"Family members should be phenotype-positive (must be high- or moderatespecificity phenotype; see phenotype table), genotype-negative 1st, 2nd, or 3rd degree relatives of the proband.",General recommendation
124
+ DICER1 (HGNC:17098),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
125
+ DICER1 (HGNC:17098),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
126
+ DICER1 (HGNC:17098),BP2,Supporting,"≥1 observation in trans with P/LP DICER1 variant or ≥3
127
+ observations in cis or phase unknown with 2+ different
128
+ P/LP DICER1 variants.",Disease-specific
129
+ DICER1 (HGNC:17098),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
130
+ DICER1 (HGNC:17098),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
131
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
132
+ DICER1 (HGNC:17098),BP4,Supporting,"For missense variants, REVEL score < 0.50 and agreement in splicing predictors that no splicing effects are predicted. For synonymous/intronic/non-coding variants concordance of MaxEntScan and SpliceAI.",Disease-specific
133
+ DICER1 (HGNC:17098),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
134
+ DICER1 (HGNC:17098),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
135
+ DICER1 (HGNC:17098),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
136
+ DICER1 (HGNC:17098),BP7,Supporting,"Silent variant
137
+ OR Intronic variant at or beyond +7 to -21 positions
138
+ OR Other intronic or non-coding variant if the variant is the reference nucleotide in ≥1 primate and/or ≥4 mammalian species.
139
+ Caveat: Variant must meet BP4 to apply BP7",General recommendation
VCI/parsing_csr_criteria/version_csv_individual/ClinGenDICER1andmiRNA-ProcessingGeneExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforDICER1Version1_version=1.0.0.csv ADDED
@@ -0,0 +1,212 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ DICER1 (HGNC:17098),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ DICER1 (HGNC:17098),PVS1,Very Strong,"Follow SVI guidance, using DICER1-specific information.
9
+ Per the PVS1 workflow guidance provided in Tayoun et al. 2018 (PMID 30192042), the following will apply:
10
+
11
+
12
+
13
+
14
+ Nonsense or frameshift variants:
15
+
16
+
17
+ PVS1 applies to variants predicted to result in nonsense-mediated decay (NMD); the predicted NMD cutoff for DICER1 occurs at p.Pro1850.
18
+
19
+
20
+ PVS1_Moderate applies to variants resulting in protein truncation 3’ of this cutoff
21
+
22
+
23
+ Canonical splice variants (+/- 1,2 intronic positions): PVS1 applies with the following exceptions:
24
+
25
+
26
+ Exon 10 SDS/SAS: PVS1_Strong (in-frame but exon includes >10% protein)
27
+
28
+
29
+ Exons 5, 15, 18, 22 SDS/SAS: PVS1_Moderate (in-frame and each <10% of protein)
30
+
31
+
32
+ Exon 27 SAS: PVS1_Moderate (final exon)
33
+
34
+
35
+ Exon 1: no criteria (non-coding)
36
+
37
+
38
+ Variants that disrupt the translation start site (p.M1?): no criteria applied given p.M1 is not highly conserved, there are three in-frame possible alternate start codons (p.Met11, p.Met17, p.Met24), and multiple lab cases of p.Met1? without
39
+ DICER1 phenotype.
40
+ SDS = splice donor site; SAS = splice acceptor site.
41
+ Refer to PS3 weight guidelines when a variant meets criterion for application of both PVS1 and PS3.
42
+ A disease-specific PVS1 decision tree incorporating the above bullets is also included at the end of this document as an additional curation tool.","Disease-specific,General recommendation"
43
+ DICER1 (HGNC:17098),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
44
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
45
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
46
+ DICER1 (HGNC:17098),PS1,Strong,"For same AA change, must confirm there is no difference in splicing using RNA data or in-silico modeling data (concordance of MaxEntScan and SpliceAI). For non-canonical intronic splicing variants at same nucleotide should have equal or worse splicing impact.
47
+ This rule code can only be used to compare variants asserted as pathogenic by the ClinGen DICER1 VCEP. Likely pathogenic changes do not apply.",General recommendation
48
+ DICER1 (HGNC:17098),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
49
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
50
+ DICER1 (HGNC:17098),PS2,Very Strong,"≥4 de novo points.
51
+ De novo points should be tallied using the simplified table for tallying
52
+ proband points and used to determine the applied strength of PS2, consistent
53
+ with SVI guidance. To avoid redundancy and increase consistency, the EP has
54
+ opted to drop PM6 and exclusively use PS2 for de novo evidence.",Strength
55
+ DICER1 (HGNC:17098),PS2,Strong,"≥2 but less than 4 de novo points.
56
+ De novo points should be tallied using the simplified table for tallying
57
+ proband points and used to determine the applied strength of PS2, consistent
58
+ with SVI guidance. To avoid redundancy and increase consistency, the EP has
59
+ opted to drop PM6 and exclusively use PS2 for de novo evidence.",Strength
60
+ DICER1 (HGNC:17098),PS2,Moderate,"≥1 but less than 2 de novo points.
61
+ De novo points should be tallied using the simplified table for tallying proband points and used to determine the applied strength of PS2, consistent with SVI guidance. To avoid redundancy and increase consistency, the EP has opted to drop PM6 and exclusively use PS2 for de novo evidence.",General recommendation
62
+ DICER1 (HGNC:17098),PS2,Supporting,"≥0.5 but less than 1 de novo points.
63
+ De novo points should be tallied using the simplified table for tallying
64
+ proband points and used to determine the applied strength of PS2, consistent with SVI guidance. To avoid redundancy and increase consistency, the EP has opted to drop PM6 and exclusively use PS2 for de novo evidence.",Strength
65
+ DICER1 (HGNC:17098),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
66
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
67
+ DICER1 (HGNC:17098),PS3,Strong,"RNA assay shows splicing impact that is out-of-frame, in-frame ≥193 residues, or in-frame with RNase IIIb disruption.
68
+ (PS3_Moderate if PVS1_Strong is applied).
69
+ This rule should be used and weighted appropriately for variants with functional evidence of a splicing impact and/or reduced DICER1 ability to cleave pre-miRNA. Follow SVI guidance regarding control numbers for functional studies. Do not apply PS3 at any strength if PVS1 is applied at full strength.",Disease-specific
70
+ DICER1 (HGNC:17098),PS3,Moderate,"RNA assay shows in-frame splicing impact with change in protein length <193 residues AND RNase IIIb domain not disrupted.
71
+ This rule should be used and weighted appropriately for variants with functional evidence of a splicing impact and/or reduced DICER1 ability to cleave pre-miRNA. Follow SVI guidance regarding control numbers for functional studies. Do not apply PS3 at any strength if PVS1 is applied at full strength.","Disease-specific,General recommendation"
72
+ DICER1 (HGNC:17098),PS3,Supporting,"In vitro cleavage assay shows failure or severely reduced capacity to produce either 5p or 3p microRNAs from a premiRNA (positive and negative controls also performed).
73
+ This rule should be used and weighted appropriately for variants with functional evidence of a splicing impact and/or reduced DICER1 ability to cleave pre-miRNA. Follow SVI guidance regarding control numbers for functional studies. Do not apply PS3 at any strength if PVS1 is applied at full strength.","Disease-specific,Strength"
74
+ DICER1 (HGNC:17098),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
75
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
76
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
77
+ DICER1 (HGNC:17098),PS4,Strong,"≥4 phenotype points.
78
+ Unrelated probands may contribute up to 1 point each based on phenotype (see Tables 2 & 3 in ruleset)
79
+ Caveats:
80
+
81
+
82
+
83
+
84
+ Do not apply PS4 if variant meets BA1/BS1 criteria.
85
+
86
+
87
+ Do not apply points for a phenotype in an individual with a likely pathogenic germline variant in a second gene that could have reasonably contributed to the phenotype (e.g. Wilms tumor in an individual with a P/LP WT1 variant).
88
+
89
+
90
+ Do not apply points for a proband whose tumor sequencing is consistent with a likely sporadic event (i.e. sequencing reveals a somatic, VCEPcurated, non-hotspot, likely pathogenic DICER1 variant in addition to a somatic hotspot variant and the germline variant under assessment). Of note, DICER1 tumors that consistently or occasionally follow a classical 2- hit hypothesis (i.e. LOF of both alleles) are exempt from this caveat. For example, identification of a somatic pathogenic non-hotspot DICER1 variant in pineoblastoma (PMID: 25022261), pituitary blastoma (PMID: 24839956), and lung cysts or cystic nephroma lacking mesenchymal cells (PMIDs: 25500911, 25978641) should not exclude the proband from PS4.",General recommendation
91
+ DICER1 (HGNC:17098),PS4,Moderate,"2 – 3.5 phenotype points.
92
+ Unrelated probands may contribute up to 1 point each based on phenotype (see Tables 2 & 3 in ruleset)
93
+ Caveats:
94
+
95
+
96
+
97
+
98
+ Do not apply PS4 if variant meets BA1/BS1 criteria.
99
+
100
+
101
+ Do not apply points for a phenotype in an individual with a likely pathogenic germline variant in a second gene that could have reasonably contributed to the phenotype (e.g. Wilms tumor in an individual with a P/LP WT1 variant).
102
+
103
+
104
+ Do not apply points for a proband whose tumor sequencing is consistent with a likely sporadic event (i.e. sequencing reveals a somatic, VCEPcurated, non-hotspot, likely pathogenic DICER1 variant in addition to a somatic hotspot variant and the germline variant under assessment). Of note, DICER1 tumors that consistently or occasionally follow a classical 2- hit hypothesis (i.e. LOF of both alleles) are exempt from this caveat. For example, identification of a somatic pathogenic non-hotspot DICER1 variant in pineoblastoma (PMID: 25022261), pituitary blastoma (PMID: 24839956), and lung cysts or cystic nephroma lacking mesenchymal cells (PMIDs: 25500911, 25978641) should not exclude the proband from PS4.",Strength
105
+ DICER1 (HGNC:17098),PS4,Supporting,"1 – 1.5 phenotype points.
106
+ Unrelated probands may contribute up to 1 point each based on phenotype (see Tables 2 & 3 in ruleset)
107
+ Caveats:
108
+
109
+
110
+
111
+
112
+ Do not apply PS4 if variant meets BA1/BS1 criteria.
113
+
114
+
115
+ Do not apply points for a phenotype in an individual with a likely pathogenic germline variant in a second gene that could have reasonably contributed to the phenotype (e.g. Wilms tumor in an individual with a P/LP WT1 variant).
116
+
117
+
118
+ Do not apply points for a proband whose tumor sequencing is consistent with a likely sporadic event (i.e. sequencing reveals a somatic, VCEPcurated, non-hotspot, likely pathogenic DICER1 variant in addition to a somatic hotspot variant and the germline variant under assessment). Of note, DICER1 tumors that consistently or occasionally follow a classical 2- hit hypothesis (i.e. LOF of both alleles) are exempt from this caveat. For example, identification of a somatic pathogenic non-hotspot DICER1 variant in pineoblastoma (PMID: 25022261), pituitary blastoma (PMID: 24839956), and lung cysts or cystic nephroma lacking mesenchymal cells (PMIDs: 25500911, 25978641) should not exclude the proband from PS4.",Strength
119
+ DICER1 (HGNC:17098),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
120
+ DICER1 (HGNC:17098),PM1,Moderate,"Residues affecting metal ion-binding: codons p.S1344, p.E1705, p.D1709, p.D1713, p.G1809, p.D1810, p.E1813",Disease-specific
121
+ DICER1 (HGNC:17098),PM1,Supporting,"Residues in the RNase IIIb domain (p.Y1682 – p.S1846),
122
+ besides the metal ion-binding residues (see PM1).",Strength
123
+ DICER1 (HGNC:17098),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
124
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
125
+ DICER1 (HGNC:17098),PM2,Supporting,Allele frequency <0.000005 across gnomAD (non-cancer) with no more than one allele in any subpopulation and at least 20x coverage.,"Disease-specific,Strength"
126
+ DICER1 (HGNC:17098),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
127
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
128
+ DICER1 (HGNC:17098),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
129
+ DICER1 (HGNC:17098),PM4,Moderate,In-frame indels with a residue within the RNase IIIb domain (p.Y1682 – p.S1846).,Disease-specific
130
+ DICER1 (HGNC:17098),PM4,Supporting,In-frame indels outside of the RNase IIIb domain (p.Y1682 – p.S1846) and repeat regions (p.D606-p.D609; p.E1418-p.E1420; p.E1422-p.E1425).,Strength
131
+ DICER1 (HGNC:17098),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
132
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
133
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
134
+ DICER1 (HGNC:17098),PM5,Moderate,"Missense variant under evaluation should have equal or worse Grantham score. Splicing should be
135
+ ruled out with either RNA data or agreement in splicing predictors (MaxEntScan and SpliceAI) that show no splicing effects (either an increase in the canonical
136
+ splice site score or a decrease in the canonical splice site score by no more than 10% AND no putative cryptic splice sites are created).
137
+ The other variant must be interpreted as pathogenic by the ClinGen DICER1 VCEP. Likely pathogenic changes do not apply. This rule cannot be applied in combination with
138
+ PM1 or PS1.",General recommendation
139
+ DICER1 (HGNC:17098),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
140
+ DICER1 (HGNC:17098),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
141
+ Note: May be used as stronger evidence with increasing segregation data.",
142
+ DICER1 (HGNC:17098),PP1,Strong,"≥7 meioses across ≥2 families.
143
+ Phenotype-positive individuals should have high, moderate, or low-specificity phenotypes (see phenotype table). (Caveat: segregation with a single low-specificity phenotype across multiple individuals (e.g. familial Wilms tumor) does not fulfill PP1.)
144
+ Do not apply PP1 if variant meets BA1/BS1 criteria.",Strength
145
+ DICER1 (HGNC:17098),PP1,Moderate,"5 – 6 meioses across ≥1 family.
146
+ Phenotype-positive individuals should have high, moderate, or low-specificity phenotypes (see phenotype table). (Caveat: segregation with a single low-specificity
147
+ phenotype across multiple individuals (e.g. familial Wilms tumor) does not fulfill PP1.)
148
+ Do not apply PP1 if variant meets BA1/BS1 criteria.",Strength
149
+ DICER1 (HGNC:17098),PP1,Supporting,"3 – 4 meioses across ≥1 family.
150
+ Phenotype-positive individuals should have high, moderate, or low-specificity phenotypes (see phenotype table). (Caveat: segregation with a single low-specificity phenotype across multiple individuals (e.g. familial Wilms tumor) does not fulfill PP1.)
151
+ Do not apply PP1 if variant meets BA1/BS1 criteria.",General recommendation
152
+ DICER1 (HGNC:17098),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
153
+ DICER1 (HGNC:17098),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
154
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
155
+ DICER1 (HGNC:17098),PP3,Supporting,"For missense variants, REVEL score ≥ 0.75 OR agreement in splicing predictors predict splicing effects. For splicing variants, concordance of MaxEntScan and SpliceAI.
156
+ PP3 may be applied for intronic variants located in reference to exons at +3 to +5 for donor splice sites or -3 to -5 for acceptor splice sites (PMID: 27313609) and have a predicted decrease in the score of the canonical splice site by at least 75%, regardless of the predicted creation or presence of a putative cryptic splice site. Must have concordance of both MaxEntScan and SpliceAI.",Disease-specific
157
+ DICER1 (HGNC:17098),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
158
+ DICER1 (HGNC:17098),PP4,Supporting,"Somatic tumor testing identifies somatic hotspot second hit and no additional somatic LOF variants.
159
+ Tumor testing (PMID: 30311369) of a neoplasm with known DICER1 association in a proband who carries the germline variant under evaluation reveals the
160
+ following:
161
+
162
+
163
+
164
+
165
+ A previously reported somatic second hit of DICER1 in an RNase IIIb-disrupting “hotspot” codon (p.S1344, p.E1705, p.D1709, p.D1713, p.G1809, p.D1810, or
166
+ p.E1813) AND
167
+
168
+
169
+ Retention of the germline DICER1 variant under evaluation.
170
+ PP4 is NOT applicable if:
171
+
172
+
173
+ The germline variant is a missense variant in one of the seven RNase IIIb “hotspot” codons (see PM1), OR
174
+
175
+
176
+ Somatic sequencing reveals additional DICER1 non-hotspot variants (could be consistent with sporadic tumorigenesis).",Disease-specific
177
+ DICER1 (HGNC:17098),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
178
+ DICER1 (HGNC:17098),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
179
+ DICER1 (HGNC:17098),BA1,Stand Alone,"Frequency >0.003 (0.3%) in gnomAD (non-cancer) subpopulations. Subpopulations must have >2,000 alleles tested and a minimum of 5 alleles present.",Disease-specific
180
+ DICER1 (HGNC:17098),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
181
+ DICER1 (HGNC:17098),BS1,Strong,"Frequency >0.0003 (0.03%) in gnomAD (non-cancer) subpopulations. Subpopulations must have >2,000 alleles tested and a minimum of 5 alleles present.",Disease-specific
182
+ DICER1 (HGNC:17098),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
183
+ DICER1 (HGNC:17098),BS2,Strong,"40+ unrelated females from a single source are tumor-free through age 50 (caveat: ratio of BS2-eligible females to PS4-eligible probands must be ≥ 40:1)
184
+ OR 2+ observations of homozygosity in healthy individuals
185
+ OR 1+ observation(s) of homozygosity in a healthy individual with status confirmed by parental testing.",Disease-specific
186
+ DICER1 (HGNC:17098),BS2,Supporting,"10+ unrelated females from a single source are tumor-free through age 50 (caveat: ratio of BS2-eligible females to PS4-eligible probands must be ≥ 10:1)
187
+ OR 2+ observations of homozygosity in individuals lacking clinical information",Disease-specific
188
+ DICER1 (HGNC:17098),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
189
+ DICER1 (HGNC:17098),BS3,Strong,"This rule should be used and weighted appropriately for variants with functional evidence of no splicing impact and/or no reduced DICER1 ability to cleave pre-miRNA. Follow SVI guidance regarding control numbers for functional studies.
190
+ For intronic or synonymous variants, no splicing impact observed via RNA assay. (Should be observed more than once.)",Disease-specific
191
+ DICER1 (HGNC:17098),BS3,Supporting,"This rule should be used and weighted appropriately for variants with functional evidence of no splicing impact and/or no reduced DICER1 ability to cleave pre-miRNA. Follow SVI guidance regarding control numbers for functional studies.
192
+ An in vitro cleavage assay must demonstrate the variant produces both 5p and 3p microRNAs from a pre-miRNA (positive and negative controls also performed). An example of an appropriate assay to which criteria could be applied is Wu et al. 2018 (PMID: 28862265).",Disease-specific
193
+ DICER1 (HGNC:17098),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
194
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
195
+ DICER1 (HGNC:17098),BS4,Strong,"Family members should be phenotype-positive (must be high- or moderatespecificity phenotype; see phenotype table), genotype-negative 1st, 2nd, or 3rd degree relatives of the proband.",General recommendation
196
+ DICER1 (HGNC:17098),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
197
+ DICER1 (HGNC:17098),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
198
+ DICER1 (HGNC:17098),BP2,Supporting,"≥1 observation in trans with P/LP DICER1 variant or ≥3
199
+ observations in cis or phase unknown with 2+ different
200
+ P/LP DICER1 variants.",Disease-specific
201
+ DICER1 (HGNC:17098),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
202
+ DICER1 (HGNC:17098),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
203
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
204
+ DICER1 (HGNC:17098),BP4,Supporting,"For missense variants, REVEL score < 0.50 and agreement in splicing predictors that no splicing effects are predicted. For synonymous/intronic/non-coding variants concordance of MaxEntScan and SpliceAI.
205
+ MaxEntScan and SpliceAI should predict either no change in splicing, an increase in canonical splice site score, or a decrease of the canonical splice site score by no more than 10% AND no putative cryptic splice sites are created.",Disease-specific
206
+ DICER1 (HGNC:17098),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,NA
207
+ DICER1 (HGNC:17098),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
208
+ DICER1 (HGNC:17098),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
209
+ DICER1 (HGNC:17098),BP7,Supporting,"Silent variant
210
+ OR Intronic variant at or beyond +7 to -21 positions
211
+ OR Other intronic or non-coding variant if the variant is the reference nucleotide in ≥1 primate and/or ≥4 mammalian species.
212
+ Caveat: Variant must meet BP4 to apply BP7",
VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA1Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,373 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ BRCA1 (HGNC:1100),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ BRCA1 (HGNC:1100),PVS1,Very Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
9
+
10
+
11
+ Well-established
12
+ in vitro
13
+ or
14
+ in vivo
15
+ functional studies supportive of a damaging effect
16
+ as measured by effect on mRNA transcript profile (mRNA assay only).
17
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
18
+ BRCA1 (HGNC:1100),PVS1,Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
19
+
20
+
21
+ Well-established
22
+ in vitro
23
+ or
24
+ in vivo
25
+ functional studies supportive of a damaging effect
26
+ as measured by effect on mRNA transcript profile (mRNA assay only).
27
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
28
+ BRCA1 (HGNC:1100),PVS1,Moderate,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
29
+
30
+
31
+ Well-established
32
+ in vitro
33
+ or
34
+ in vivo
35
+ functional studies supportive of a damaging effect
36
+ as measured by effect on mRNA transcript profile (mRNA assay only).
37
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
38
+ BRCA1 (HGNC:1100),PVS1,Supporting,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
39
+
40
+
41
+ Well-established
42
+ in vitro
43
+ or
44
+ in vivo
45
+ functional studies supportive of a damaging effect
46
+ as measured by effect on mRNA transcript profile (mRNA assay only).
47
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
48
+ BRCA1 (HGNC:1100),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
49
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
50
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
51
+ BRCA1 (HGNC:1100),PS1,Strong,"Apply
52
+ PS1
53
+ , for predicted
54
+ missense
55
+ substitutions, where a previously classified
56
+ pathogenic
57
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
58
+
59
+
60
+ Apply
61
+ PS1
62
+ , for exonic and intronic variants with same predicted impact on
63
+ splicing
64
+ , as a previously classified
65
+ pathogenic
66
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant. 
67
+
68
+
69
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
70
+ BRCA1 (HGNC:1100),PS1,Moderate,"Apply
71
+ PS1_Moderate
72
+ , for predicted
73
+ missense
74
+ substitutions, where previously classified
75
+ likely pathogenic
76
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
77
+
78
+
79
+ Apply
80
+ PS1_Moderate
81
+ , for exonic and intronic variants with same predicted impact on
82
+ splicing
83
+ , as a previously classified
84
+ (likely) pathogenic
85
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
86
+
87
+
88
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
89
+ BRCA1 (HGNC:1100),PS1,Supporting,"Apply
90
+ PS1_Supporting
91
+ , for exonic and intronic variants with same predicted impact on
92
+ splicing,
93
+ as a previously classified
94
+ (likely) pathogenic
95
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
96
+
97
+
98
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
99
+ BRCA1 (HGNC:1100),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
100
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
101
+ BRCA1 (HGNC:1100),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
102
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
103
+ BRCA1 (HGNC:1100),PS3,Strong,"Well-established in vitro or in vivo functional studies supportive of a damaging effect. Apply PS3 for assays measuring effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
104
+
105
+
106
+ Well-established
107
+ in vitro
108
+ or
109
+ in vivo
110
+ functional studies supportive of a damaging effect
111
+ as measured by effect on mRNA transcript profile (mRNA assay only).
112
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",General recommendation
113
+ BRCA1 (HGNC:1100),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
114
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
115
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
116
+ BRCA1 (HGNC:1100),PS4,Strong,The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls. Case-control studies; p-value ≤0.05 and OR ≥4 (lower confidence interval excludes 2.0). See Appendix F for details.,"Clarification,Gene-specific"
117
+ BRCA1 (HGNC:1100),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
118
+ BRCA1 (HGNC:1100),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
119
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
120
+ BRCA1 (HGNC:1100),PM2,Supporting,"Absent from controls in an outbred population, from gnomAD v2.1 (non-cancer, exome only subset) and gnomAD v3.1 (non-cancer). Region around the variant must have an average read depth ≥25. See Appendix G for details.",Gene-specific
121
+ BRCA1 (HGNC:1100),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
122
+ Note: This requires testing of parents (or offspring) to determine phase.",
123
+ BRCA1 (HGNC:1100),PM3,Strong,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
124
+
125
+
126
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
127
+ ≤5yr
128
+ .
129
+
130
+
131
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
132
+
133
+
134
+ See
135
+ Specifications Table 6
136
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
137
+
138
+
139
+ PM3_Strong = ≥4 points",Gene-specific
140
+ BRCA1 (HGNC:1100),PM3,Moderate,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
141
+
142
+
143
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
144
+ ≤5yr
145
+ .
146
+
147
+
148
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
149
+
150
+
151
+ See
152
+ Specifications Table 6
153
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
154
+
155
+
156
+ PM3 = 2 points",Gene-specific
157
+ BRCA1 (HGNC:1100),PM3,Supporting,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
158
+
159
+
160
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
161
+ ≤5yr
162
+ .
163
+
164
+
165
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
166
+
167
+
168
+ See
169
+ Specifications Table 6
170
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
171
+
172
+
173
+ PM3_Supporting = 1 point",Gene-specific
174
+ BRCA1 (HGNC:1100),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,NA
175
+ BRCA1 (HGNC:1100),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
176
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
177
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
178
+ BRCA1 (HGNC:1100),PM5,Strong,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
179
+ BRCA1 (HGNC:1100),PM5,Moderate,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
180
+ BRCA1 (HGNC:1100),PM5,Supporting,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
181
+ BRCA1 (HGNC:1100),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
182
+ BRCA1 (HGNC:1100),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
183
+ Note: May be used as stronger evidence with increasing segregation data.",
184
+ BRCA1 (HGNC:1100),PP1,Strong,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
185
+
186
+
187
+ Apply weight as per Bayes Score:
188
+
189
+
190
+ PP1_Strong – LR>18.7:1
191
+
192
+
193
+ PP1_Very Strong – LR>350:1",Gene-specific
194
+ BRCA1 (HGNC:1100),PP1,Moderate,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
195
+
196
+
197
+ Apply weight as per Bayes Score:
198
+
199
+
200
+ PP1_Moderate – LR>4.3:1",Gene-specific
201
+ BRCA1 (HGNC:1100),PP1,Supporting,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
202
+
203
+
204
+ Apply weight as per Bayes Score:
205
+
206
+
207
+ PP1 - LR >2.08:1",Gene-specific
208
+ BRCA1 (HGNC:1100),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
209
+ BRCA1 (HGNC:1100),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
210
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
211
+ BRCA1 (HGNC:1100),PP3,Supporting,"Apply PP3 for missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain and predicted impact via protein change (BayesDel no-AF score ≥0.28). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857.
212
+
213
+
214
+ Apply PP3 for predicted splicing (SpliceAI ≥0.2) for silent, missense/in-frame (irrespective of location in clinically important functional domain) and for intronic variants outside of donor and acceptor 1,2 sites.
215
+
216
+
217
+ See Specifications Figure1A and Appendix J for details.",Gene-specific
218
+ BRCA1 (HGNC:1100),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
219
+ BRCA1 (HGNC:1100),PP4,Strong,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
220
+
221
+
222
+ PP4_Strong – LR>18.7:1
223
+
224
+
225
+ PP4_Very Strong – LR>350:1
226
+
227
+
228
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
229
+
230
+
231
+ See Specifications Table7 and Appendix B for details.",Gene-specific
232
+ BRCA1 (HGNC:1100),PP4,Moderate,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
233
+
234
+
235
+ PP4_Moderate – LR>4.3:1
236
+
237
+
238
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
239
+
240
+
241
+ See Specifications Table7 and Appendix B for details.",Gene-specific
242
+ BRCA1 (HGNC:1100),PP4,Supporting,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
243
+
244
+
245
+ PP4 - LR >2.08:1 
246
+
247
+
248
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
249
+
250
+
251
+ See Specifications Table7 and Appendix B for details.",Gene-specific
252
+ BRCA1 (HGNC:1100),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
253
+ BRCA1 (HGNC:1100),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
254
+ BRCA1 (HGNC:1100),BA1,Stand Alone,"Filter allele frequency (FAF) is above 0.1% (FAF > 0.001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
255
+ BRCA1 (HGNC:1100),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
256
+ BRCA1 (HGNC:1100),BS1,Strong,"Filter allele frequency (FAF) is above 0.01% (FAF > 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
257
+ BRCA1 (HGNC:1100),BS1,Supporting,"Filter allele frequency (FAF) is above 0.002% (FAF > 0.00002) and less than or equal to 0.01% (FAF ≤ 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
258
+ BRCA1 (HGNC:1100),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
259
+ BRCA1 (HGNC:1100),BS2,Strong,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
260
+ Specifications Table 8
261
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
262
+
263
+
264
+ BS2 = ≥ 4 points",Gene-specific
265
+ BRCA1 (HGNC:1100),BS2,Moderate,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
266
+ Specifications Table 8
267
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
268
+
269
+
270
+ BS2_Moderate = 2 points",Gene-specific
271
+ BRCA1 (HGNC:1100),BS2,Supporting,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
272
+ Specifications Table 8
273
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
274
+
275
+
276
+ BS2_Supporting = 1 points",Gene-specific
277
+ BRCA1 (HGNC:1100),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
278
+ BRCA1 (HGNC:1100),BS3,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function. Assay measures effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
279
+
280
+
281
+ Well-established
282
+ in vitro
283
+ or
284
+ in vivo
285
+ functional studies supportive of no damaging effect
286
+ as measured by effect on mRNA transcript profile (mRNA assay only).
287
+ Apply as BP7 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
288
+ BRCA1 (HGNC:1100),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
289
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
290
+ BRCA1 (HGNC:1100),BS4,Strong,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
291
+
292
+
293
+ Apply weight as per Bayes Score:
294
+
295
+
296
+ BS4 - LR <0.05:1
297
+
298
+
299
+ BS4_VeryStrong – LR <0.00285:1",Gene-specific
300
+ BRCA1 (HGNC:1100),BS4,Moderate,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
301
+
302
+
303
+ Apply weight as per Bayes Score:
304
+
305
+
306
+ BS4_Moderate - LR <0.23:1",Gene-specific
307
+ BRCA1 (HGNC:1100),BS4,Supporting,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
308
+
309
+
310
+ Apply weight as per Bayes Score:
311
+
312
+
313
+ BS4_Supporting  - LR 0.23-0.48:1",Gene-specific
314
+ BRCA1 (HGNC:1100),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,
315
+ BRCA1 (HGNC:1100),BP1,Strong,"Apply BP1_Strong
316
+ for silent substitution, missense or in-frame insertion, deletion or delins variants outside a (potentially) clinically important functional domain AND no splicing predicted (SpliceAI ≤0.1). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","Gene-specific,Strength"
317
+ BRCA1 (HGNC:1100),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,NA
318
+ BRCA1 (HGNC:1100),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
319
+ BRCA1 (HGNC:1100),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
320
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
321
+ BRCA1 (HGNC:1100),BP4,Supporting,"Missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain, and no predicted impact via protein change or splicing (BayesDel no-AF score ≤ 0.15 AND SpliceAI ≤0.1).
322
+
323
+
324
+ Silent variant inside a (potentially) clinically important functional domain, if no predicted impact via splicing (SpliceAI ≤0.1).
325
+
326
+
327
+ Intronic variants outside of the native donor and acceptor splice sites (i.e. not +/- 1,2 positions) AND no predicted impact via splicing (SpliceAI ≤0.1).
328
+
329
+
330
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","Clarification,Gene-specific"
331
+ BRCA1 (HGNC:1100),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
332
+ BRCA1 (HGNC:1100),BP5,Strong,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
333
+
334
+
335
+ BP5_VeryStrong – LR <0.00285:1
336
+
337
+
338
+ BP5_Strong  - LR <0.05:1
339
+
340
+
341
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
342
+ BRCA1 (HGNC:1100),BP5,Moderate,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
343
+
344
+
345
+ BP5_Moderate - LR <0.23:1
346
+
347
+
348
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
349
+ BRCA1 (HGNC:1100),BP5,Supporting,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
350
+
351
+
352
+ BP5 - LR 0.23-0.48:1
353
+
354
+
355
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
356
+ BRCA1 (HGNC:1100),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
357
+ BRCA1 (HGNC:1100),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
358
+ BRCA1 (HGNC:1100),BP7,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function
359
+ as measured by effect on mRNA transcript profile – mRNA assay only.
360
+ Apply as BP7 (RNA) for intronic and silent variants, as well as missense/in-frame variants located outside a (potentially) clinically important functional domain. See Specifications Figure1B and Appendix E for details.
361
+
362
+
363
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","General recommendation,Gene-specific"
364
+ BRCA1 (HGNC:1100),BP7,Supporting,"Silent variant inside a (potentially) clinically important functional domain, IF BP4 met.
365
+
366
+
367
+ Intronic variants located outside conserved donor or acceptor motif positions (at or beyond positions +7/-21) IF BP4 met.
368
+
369
+
370
+ See Specifications Figure1A and Appendix J for additional details.
371
+
372
+
373
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","Clarification,General recommendation"
VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA1Version1.1.0_version=1.1.0.csv ADDED
@@ -0,0 +1,373 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ BRCA1 (HGNC:1100),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ BRCA1 (HGNC:1100),PVS1,Very Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
9
+
10
+
11
+ Well-established
12
+ in vitro
13
+ or
14
+ in vivo
15
+ functional studies supportive of a damaging effect
16
+ as measured by effect on mRNA transcript profile (mRNA assay only).
17
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
18
+ BRCA1 (HGNC:1100),PVS1,Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
19
+
20
+
21
+ Well-established
22
+ in vitro
23
+ or
24
+ in vivo
25
+ functional studies supportive of a damaging effect
26
+ as measured by effect on mRNA transcript profile (mRNA assay only).
27
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
28
+ BRCA1 (HGNC:1100),PVS1,Moderate,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
29
+
30
+
31
+ Well-established
32
+ in vitro
33
+ or
34
+ in vivo
35
+ functional studies supportive of a damaging effect
36
+ as measured by effect on mRNA transcript profile (mRNA assay only).
37
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
38
+ BRCA1 (HGNC:1100),PVS1,Supporting,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
39
+
40
+
41
+ Well-established
42
+ in vitro
43
+ or
44
+ in vivo
45
+ functional studies supportive of a damaging effect
46
+ as measured by effect on mRNA transcript profile (mRNA assay only).
47
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
48
+ BRCA1 (HGNC:1100),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
49
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
50
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
51
+ BRCA1 (HGNC:1100),PS1,Strong,"Apply
52
+ PS1
53
+ , for predicted
54
+ missense
55
+ substitutions, where a previously classified
56
+ pathogenic
57
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
58
+
59
+
60
+ Apply
61
+ PS1
62
+ , for exonic and intronic variants with same predicted impact on
63
+ splicing
64
+ , as a previously classified
65
+ pathogenic
66
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant. 
67
+
68
+
69
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
70
+ BRCA1 (HGNC:1100),PS1,Moderate,"Apply
71
+ PS1_Moderate
72
+ , for predicted
73
+ missense
74
+ substitutions, where previously classified
75
+ likely pathogenic
76
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
77
+
78
+
79
+ Apply
80
+ PS1_Moderate
81
+ , for exonic and intronic variants with same predicted impact on
82
+ splicing
83
+ , as a previously classified
84
+ (likely) pathogenic
85
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
86
+
87
+
88
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
89
+ BRCA1 (HGNC:1100),PS1,Supporting,"Apply
90
+ PS1_Supporting
91
+ , for exonic and intronic variants with same predicted impact on
92
+ splicing,
93
+ as a previously classified
94
+ (likely) pathogenic
95
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
96
+
97
+
98
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
99
+ BRCA1 (HGNC:1100),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
100
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
101
+ BRCA1 (HGNC:1100),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
102
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
103
+ BRCA1 (HGNC:1100),PS3,Strong,"Well-established in vitro or in vivo functional studies supportive of a damaging effect. Apply PS3 for assays measuring effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
104
+
105
+
106
+ Well-established
107
+ in vitro
108
+ or
109
+ in vivo
110
+ functional studies supportive of a damaging effect
111
+ as measured by effect on mRNA transcript profile (mRNA assay only).
112
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",General recommendation
113
+ BRCA1 (HGNC:1100),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
114
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
115
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
116
+ BRCA1 (HGNC:1100),PS4,Strong,The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls. Case-control studies; p-value ≤0.05 and OR ≥4 (lower confidence interval excludes 2.0). See Appendix F for details.,"Clarification,Gene-specific"
117
+ BRCA1 (HGNC:1100),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
118
+ BRCA1 (HGNC:1100),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
119
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
120
+ BRCA1 (HGNC:1100),PM2,Supporting,"Absent from controls in an outbred population, from gnomAD v2.1 (non-cancer, exome only subset) and gnomAD v3.1 (non-cancer). Region around the variant must have an average read depth ≥25. See Appendix G for details.",Gene-specific
121
+ BRCA1 (HGNC:1100),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
122
+ Note: This requires testing of parents (or offspring) to determine phase.",
123
+ BRCA1 (HGNC:1100),PM3,Strong,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
124
+
125
+
126
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
127
+ ≤5yr
128
+ .
129
+
130
+
131
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
132
+
133
+
134
+ See
135
+ Specifications Table 6
136
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
137
+
138
+
139
+ PM3_Strong = ≥4 points",Gene-specific
140
+ BRCA1 (HGNC:1100),PM3,Moderate,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
141
+
142
+
143
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
144
+ ≤5yr
145
+ .
146
+
147
+
148
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
149
+
150
+
151
+ See
152
+ Specifications Table 6
153
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
154
+
155
+
156
+ PM3 = 2 points",Gene-specific
157
+ BRCA1 (HGNC:1100),PM3,Supporting,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
158
+
159
+
160
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
161
+ ≤5yr
162
+ .
163
+
164
+
165
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
166
+
167
+
168
+ See
169
+ Specifications Table 6
170
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
171
+
172
+
173
+ PM3_Supporting = 1 point",Gene-specific
174
+ BRCA1 (HGNC:1100),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,NA
175
+ BRCA1 (HGNC:1100),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
176
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
177
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
178
+ BRCA1 (HGNC:1100),PM5,Strong,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
179
+ BRCA1 (HGNC:1100),PM5,Moderate,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
180
+ BRCA1 (HGNC:1100),PM5,Supporting,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
181
+ BRCA1 (HGNC:1100),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
182
+ BRCA1 (HGNC:1100),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
183
+ Note: May be used as stronger evidence with increasing segregation data.",
184
+ BRCA1 (HGNC:1100),PP1,Strong,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
185
+
186
+
187
+ Apply weight as per Bayes Score:
188
+
189
+
190
+ PP1_Strong – LR>18.7:1
191
+
192
+
193
+ PP1_Very Strong – LR>350:1",Gene-specific
194
+ BRCA1 (HGNC:1100),PP1,Moderate,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
195
+
196
+
197
+ Apply weight as per Bayes Score:
198
+
199
+
200
+ PP1_Moderate – LR>4.3:1",Gene-specific
201
+ BRCA1 (HGNC:1100),PP1,Supporting,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
202
+
203
+
204
+ Apply weight as per Bayes Score:
205
+
206
+
207
+ PP1 - LR >2.08:1",Gene-specific
208
+ BRCA1 (HGNC:1100),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
209
+ BRCA1 (HGNC:1100),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
210
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
211
+ BRCA1 (HGNC:1100),PP3,Supporting,"Apply PP3 for missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain and predicted impact via protein change (BayesDel no-AF score ≥0.28). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857.
212
+
213
+
214
+ Apply PP3 for predicted splicing (SpliceAI ≥0.2) for silent, missense/in-frame (irrespective of location in clinically important functional domain) and for intronic variants outside of donor and acceptor 1,2 sites.
215
+
216
+
217
+ See Specifications Figure1A and Appendix J for details.",Gene-specific
218
+ BRCA1 (HGNC:1100),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
219
+ BRCA1 (HGNC:1100),PP4,Strong,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
220
+
221
+
222
+ PP4_Strong – LR>18.7:1
223
+
224
+
225
+ PP4_Very Strong – LR>350:1
226
+
227
+
228
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
229
+
230
+
231
+ See Specifications Table7 and Appendix B for details.",Gene-specific
232
+ BRCA1 (HGNC:1100),PP4,Moderate,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
233
+
234
+
235
+ PP4_Moderate – LR>4.3:1
236
+
237
+
238
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
239
+
240
+
241
+ See Specifications Table7 and Appendix B for details.",Gene-specific
242
+ BRCA1 (HGNC:1100),PP4,Supporting,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
243
+
244
+
245
+ PP4 - LR >2.08:1 
246
+
247
+
248
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
249
+
250
+
251
+ See Specifications Table7 and Appendix B for details.",Gene-specific
252
+ BRCA1 (HGNC:1100),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
253
+ BRCA1 (HGNC:1100),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
254
+ BRCA1 (HGNC:1100),BA1,Stand Alone,"Filter allele frequency (FAF) is above 0.1% (FAF > 0.001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
255
+ BRCA1 (HGNC:1100),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
256
+ BRCA1 (HGNC:1100),BS1,Strong,"Filter allele frequency (FAF) is above 0.01% (FAF > 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
257
+ BRCA1 (HGNC:1100),BS1,Supporting,"Filter allele frequency (FAF) is above 0.002% (FAF > 0.00002) and less than or equal to 0.01% (FAF ≤ 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
258
+ BRCA1 (HGNC:1100),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
259
+ BRCA1 (HGNC:1100),BS2,Strong,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
260
+ Specifications Table 8
261
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
262
+
263
+
264
+ BS2 = ≥ 4 points",Gene-specific
265
+ BRCA1 (HGNC:1100),BS2,Moderate,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
266
+ Specifications Table 8
267
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
268
+
269
+
270
+ BS2_Moderate = 2 points",Gene-specific
271
+ BRCA1 (HGNC:1100),BS2,Supporting,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
272
+ Specifications Table 8
273
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
274
+
275
+
276
+ BS2_Supporting = 1 points",Gene-specific
277
+ BRCA1 (HGNC:1100),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
278
+ BRCA1 (HGNC:1100),BS3,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function. Assay measures effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
279
+
280
+
281
+ Well-established
282
+ in vitro
283
+ or
284
+ in vivo
285
+ functional studies supportive of no damaging effect
286
+ as measured by effect on mRNA transcript profile (mRNA assay only).
287
+ Apply as BP7 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
288
+ BRCA1 (HGNC:1100),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
289
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
290
+ BRCA1 (HGNC:1100),BS4,Strong,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
291
+
292
+
293
+ Apply weight as per Bayes Score:
294
+
295
+
296
+ BS4 - LR <0.05:1
297
+
298
+
299
+ BS4_VeryStrong – LR <0.00285:1",Gene-specific
300
+ BRCA1 (HGNC:1100),BS4,Moderate,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
301
+
302
+
303
+ Apply weight as per Bayes Score:
304
+
305
+
306
+ BS4_Moderate - LR <0.23:1",Gene-specific
307
+ BRCA1 (HGNC:1100),BS4,Supporting,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
308
+
309
+
310
+ Apply weight as per Bayes Score:
311
+
312
+
313
+ BS4_Supporting  - LR 0.23-0.48:1",Gene-specific
314
+ BRCA1 (HGNC:1100),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,
315
+ BRCA1 (HGNC:1100),BP1,Strong,"Apply BP1_Strong
316
+ for silent substitution, missense or in-frame insertion, deletion or delins variants outside a (potentially) clinically important functional domain AND no splicing predicted (SpliceAI ≤0.1). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","Gene-specific,Strength"
317
+ BRCA1 (HGNC:1100),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,NA
318
+ BRCA1 (HGNC:1100),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
319
+ BRCA1 (HGNC:1100),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
320
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
321
+ BRCA1 (HGNC:1100),BP4,Supporting,"Missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain, and no predicted impact via protein change or splicing (BayesDel no-AF score ≤ 0.15 AND SpliceAI ≤0.1).
322
+
323
+
324
+ Silent variant inside a (potentially) clinically important functional domain, if no predicted impact via splicing (SpliceAI ≤0.1).
325
+
326
+
327
+ Intronic variants outside of the native donor and acceptor splice sites (i.e. not +/- 1,2 positions) AND no predicted impact via splicing (SpliceAI ≤0.1).
328
+
329
+
330
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","Clarification,Gene-specific"
331
+ BRCA1 (HGNC:1100),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
332
+ BRCA1 (HGNC:1100),BP5,Strong,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
333
+
334
+
335
+ BP5_VeryStrong – LR <0.00285:1
336
+
337
+
338
+ BP5_Strong  - LR <0.05:1
339
+
340
+
341
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
342
+ BRCA1 (HGNC:1100),BP5,Moderate,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
343
+
344
+
345
+ BP5_Moderate - LR <0.23:1
346
+
347
+
348
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
349
+ BRCA1 (HGNC:1100),BP5,Supporting,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
350
+
351
+
352
+ BP5 - LR 0.23-0.48:1
353
+
354
+
355
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
356
+ BRCA1 (HGNC:1100),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
357
+ BRCA1 (HGNC:1100),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
358
+ BRCA1 (HGNC:1100),BP7,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function
359
+ as measured by effect on mRNA transcript profile – mRNA assay only.
360
+ Apply as BP7 (RNA) for intronic and silent variants, as well as missense/in-frame variants located outside a (potentially) clinically important functional domain. See Specifications Figure1B and Appendix E for details.
361
+
362
+
363
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","General recommendation,Gene-specific"
364
+ BRCA1 (HGNC:1100),BP7,Supporting,"Silent variant inside a (potentially) clinically important functional domain, IF BP4 met.
365
+
366
+
367
+ Intronic variants located outside conserved donor or acceptor motif positions (at or beyond positions +7/-21) IF BP4 met.
368
+
369
+
370
+ See Specifications Figure1A and Appendix J for additional details.
371
+
372
+
373
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","Clarification,General recommendation"
VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA1Version1.2.0_version=1.2.0.csv ADDED
@@ -0,0 +1,364 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ BRCA1 (HGNC:1100),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ BRCA1 (HGNC:1100),PVS1,Very Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
9
+
10
+
11
+ Well-established
12
+ in vitro
13
+ or
14
+ in vivo
15
+ functional studies supportive of a damaging effect
16
+ as measured by effect on mRNA transcript profile (mRNA assay only).
17
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
18
+ BRCA1 (HGNC:1100),PVS1,Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
19
+
20
+
21
+ Well-established
22
+ in vitro
23
+ or
24
+ in vivo
25
+ functional studies supportive of a damaging effect
26
+ as measured by effect on mRNA transcript profile (mRNA assay only).
27
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
28
+ BRCA1 (HGNC:1100),PVS1,Moderate,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
29
+
30
+
31
+ Well-established
32
+ in vitro
33
+ or
34
+ in vivo
35
+ functional studies supportive of a damaging effect
36
+ as measured by effect on mRNA transcript profile (mRNA assay only).
37
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
38
+ BRCA1 (HGNC:1100),PVS1,Supporting,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
39
+
40
+
41
+ Well-established
42
+ in vitro
43
+ or
44
+ in vivo
45
+ functional studies supportive of a damaging effect
46
+ as measured by effect on mRNA transcript profile (mRNA assay only).
47
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
48
+ BRCA1 (HGNC:1100),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
49
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
50
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
51
+ BRCA1 (HGNC:1100),PS1,Strong,"Apply
52
+ PS1
53
+ , for predicted
54
+ missense
55
+ substitutions, where a previously classified
56
+ pathogenic
57
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
58
+
59
+
60
+ Apply
61
+ PS1
62
+ , for exonic and intronic variants with same predicted impact on
63
+ splicing
64
+ , as a previously classified
65
+ pathogenic
66
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant. 
67
+
68
+
69
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
70
+ BRCA1 (HGNC:1100),PS1,Moderate,"Apply
71
+ PS1_Moderate
72
+ , for predicted
73
+ missense
74
+ substitutions, where previously classified
75
+ likely pathogenic
76
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
77
+
78
+
79
+ Apply
80
+ PS1_Moderate
81
+ , for exonic and intronic variants with same predicted impact on
82
+ splicing
83
+ , as a previously classified
84
+ (likely) pathogenic
85
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
86
+
87
+
88
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
89
+ BRCA1 (HGNC:1100),PS1,Supporting,"Apply
90
+ PS1_Supporting
91
+ , for exonic and intronic variants with same predicted impact on
92
+ splicing,
93
+ as a previously classified
94
+ (likely) pathogenic
95
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
96
+
97
+
98
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
99
+ BRCA1 (HGNC:1100),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
100
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
101
+ BRCA1 (HGNC:1100),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
102
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
103
+ BRCA1 (HGNC:1100),PS3,Strong,"Well-established in vitro or in vivo functional studies supportive of a damaging effect. Apply PS3 for assays measuring effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
104
+
105
+
106
+ Well-established
107
+ in vitro
108
+ or
109
+ in vivo
110
+ functional studies supportive of a damaging effect
111
+ as measured by effect on mRNA transcript profile (mRNA assay only).
112
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",General recommendation
113
+ BRCA1 (HGNC:1100),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
114
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
115
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
116
+ BRCA1 (HGNC:1100),PS4,Strong,The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls. Case-control studies; p-value ≤0.05 and OR ≥4 (lower confidence interval excludes 2.0). See Appendix F for details.,"Clarification,Gene-specific"
117
+ BRCA1 (HGNC:1100),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
118
+ BRCA1 (HGNC:1100),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
119
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
120
+ BRCA1 (HGNC:1100),PM2,Supporting,"Absent from controls in an outbred population, from gnomAD v2.1 (non-cancer, exome only subset) and gnomAD v3.1 (non-cancer). Region around the variant must have an average read depth ≥25. See Appendix G for details.",Gene-specific
121
+ BRCA1 (HGNC:1100),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
122
+ Note: This requires testing of parents (or offspring) to determine phase.",
123
+ BRCA1 (HGNC:1100),PM3,Strong,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
124
+
125
+
126
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
127
+ ≤5yr
128
+ .
129
+
130
+
131
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
132
+
133
+
134
+ See
135
+ Specifications Table 6
136
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
137
+
138
+
139
+ PM3_Strong = ≥4 points",Gene-specific
140
+ BRCA1 (HGNC:1100),PM3,Moderate,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
141
+
142
+
143
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
144
+ ≤5yr
145
+ .
146
+
147
+
148
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
149
+
150
+
151
+ See
152
+ Specifications Table 6
153
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
154
+
155
+
156
+ PM3 = 2 points",Gene-specific
157
+ BRCA1 (HGNC:1100),PM3,Supporting,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
158
+
159
+
160
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
161
+ ≤5yr
162
+ .
163
+
164
+
165
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
166
+
167
+
168
+ See
169
+ Specifications Table 6
170
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
171
+
172
+
173
+ PM3_Supporting = 1 point",Gene-specific
174
+ BRCA1 (HGNC:1100),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,NA
175
+ BRCA1 (HGNC:1100),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
176
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
177
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
178
+ BRCA1 (HGNC:1100),PM5,Strong,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
179
+ BRCA1 (HGNC:1100),PM5,Moderate,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
180
+ BRCA1 (HGNC:1100),PM5,Supporting,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
181
+ BRCA1 (HGNC:1100),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
182
+ BRCA1 (HGNC:1100),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
183
+ Note: May be used as stronger evidence with increasing segregation data.",
184
+ BRCA1 (HGNC:1100),PP1,Strong,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
185
+
186
+
187
+ Apply weight as per Bayes Score:
188
+
189
+
190
+ PP1_Strong – LR ≥18.7:1
191
+
192
+
193
+ PP1_Very Strong – LR ≥350:1",Gene-specific
194
+ BRCA1 (HGNC:1100),PP1,Moderate,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
195
+
196
+
197
+ Apply weight as per Bayes Score:
198
+
199
+
200
+ PP1_Moderate – LR ≥4.3:1",Gene-specific
201
+ BRCA1 (HGNC:1100),PP1,Supporting,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
202
+
203
+
204
+ Apply weight as per Bayes Score:
205
+
206
+
207
+ PP1 - LR ≥2.08:1",Gene-specific
208
+ BRCA1 (HGNC:1100),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
209
+ BRCA1 (HGNC:1100),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
210
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
211
+ BRCA1 (HGNC:1100),PP3,Supporting,"Apply PP3 for missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain and predicted impact via protein change (BayesDel no-AF score ≥0.28). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857.
212
+
213
+
214
+ Apply PP3 for predicted splicing (SpliceAI ≥0.2) for silent, missense/in-frame (irrespective of location in clinically important functional domain) and for intronic variants outside of donor and acceptor 1,2 sites.
215
+
216
+
217
+ See Specifications Figure1A and Appendix J for details.",Gene-specific
218
+ BRCA1 (HGNC:1100),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
219
+ BRCA1 (HGNC:1100),PP4,Strong,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
220
+
221
+
222
+ PP4_Strong – LR ≥18.7:1
223
+
224
+
225
+ PP4_Very Strong – LR ≥350:1
226
+
227
+
228
+ See Specifications Table7 and Appendix B for details.",Gene-specific
229
+ BRCA1 (HGNC:1100),PP4,Moderate,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
230
+
231
+
232
+ PP4_Moderate – LR ≥4.3:1
233
+
234
+
235
+ See Specifications Table7 and Appendix B for details.",Gene-specific
236
+ BRCA1 (HGNC:1100),PP4,Supporting,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
237
+
238
+
239
+ PP4 - LR ≥2.08:1 
240
+
241
+
242
+ See Specifications Table7 and Appendix B for details.",Gene-specific
243
+ BRCA1 (HGNC:1100),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
244
+ BRCA1 (HGNC:1100),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
245
+ BRCA1 (HGNC:1100),BA1,Stand Alone,"Filter allele frequency (FAF) is above 0.1% (FAF > 0.001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
246
+ BRCA1 (HGNC:1100),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
247
+ BRCA1 (HGNC:1100),BS1,Strong,"Filter allele frequency (FAF) is above 0.01% (FAF > 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
248
+ BRCA1 (HGNC:1100),BS1,Supporting,"Filter allele frequency (FAF) is above 0.002% (FAF > 0.00002) and less than or equal to 0.01% (FAF ≤ 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
249
+ BRCA1 (HGNC:1100),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
250
+ BRCA1 (HGNC:1100),BS2,Strong,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
251
+ Specifications Table 8
252
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
253
+
254
+
255
+ BS2 = ≥ 4 points",Gene-specific
256
+ BRCA1 (HGNC:1100),BS2,Moderate,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
257
+ Specifications Table 8
258
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
259
+
260
+
261
+ BS2_Moderate = 2 points",Gene-specific
262
+ BRCA1 (HGNC:1100),BS2,Supporting,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
263
+ Specifications Table 8
264
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
265
+
266
+
267
+ BS2_Supporting = 1 points",Gene-specific
268
+ BRCA1 (HGNC:1100),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
269
+ BRCA1 (HGNC:1100),BS3,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function. Assay measures effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
270
+
271
+
272
+ Well-established
273
+ in vitro
274
+ or
275
+ in vivo
276
+ functional studies supportive of no damaging effect
277
+ as measured by effect on mRNA transcript profile (mRNA assay only).
278
+ Apply as BP7 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
279
+ BRCA1 (HGNC:1100),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
280
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
281
+ BRCA1 (HGNC:1100),BS4,Strong,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
282
+
283
+
284
+ Apply weight as per Bayes Score:
285
+
286
+
287
+ BS4 - LR ≤0.05:1
288
+
289
+
290
+ BS4_VeryStrong – LR ≤0.00285:1",Gene-specific
291
+ BRCA1 (HGNC:1100),BS4,Moderate,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
292
+
293
+
294
+ Apply weight as per Bayes Score:
295
+
296
+
297
+ BS4_Moderate - LR ≤0.23:1",Gene-specific
298
+ BRCA1 (HGNC:1100),BS4,Supporting,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
299
+
300
+
301
+ Apply weight as per Bayes Score:
302
+
303
+
304
+ BS4_Supporting  - LR ≤0.48:1",Gene-specific
305
+ BRCA1 (HGNC:1100),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,
306
+ BRCA1 (HGNC:1100),BP1,Strong,"Apply BP1_Strong
307
+ for silent substitution, missense or in-frame insertion, deletion or delins variants outside a (potentially) clinically important functional domain AND no splicing predicted (SpliceAI ≤0.1). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","Gene-specific,Strength"
308
+ BRCA1 (HGNC:1100),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,NA
309
+ BRCA1 (HGNC:1100),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
310
+ BRCA1 (HGNC:1100),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
311
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
312
+ BRCA1 (HGNC:1100),BP4,Supporting,"Missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain, and no predicted impact via protein change or splicing (BayesDel no-AF score ≤0.15 AND SpliceAI ≤0.1).
313
+
314
+
315
+ Silent variant inside a (potentially) clinically important functional domain, if no predicted impact via splicing (SpliceAI ≤0.1).
316
+
317
+
318
+ Intronic variants outside of the native donor and acceptor splice sites (i.e. not +/- 1,2 positions) AND no predicted impact via splicing (SpliceAI ≤0.1).
319
+
320
+
321
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","Clarification,Gene-specific"
322
+ BRCA1 (HGNC:1100),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
323
+ BRCA1 (HGNC:1100),BP5,Strong,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
324
+
325
+
326
+ BP5_VeryStrong - LR ≤0.00285:1
327
+
328
+
329
+ BP5_Strong - LR ≤0.05:1
330
+
331
+
332
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
333
+ BRCA1 (HGNC:1100),BP5,Moderate,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
334
+
335
+
336
+ BP5_Moderate - LR ≤0.23:1
337
+
338
+
339
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
340
+ BRCA1 (HGNC:1100),BP5,Supporting,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
341
+
342
+
343
+ BP5 - LR ≤0.48:1
344
+
345
+
346
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
347
+ BRCA1 (HGNC:1100),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
348
+ BRCA1 (HGNC:1100),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
349
+ BRCA1 (HGNC:1100),BP7,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function
350
+ as measured by effect on mRNA transcript profile – mRNA assay only.
351
+ Apply as BP7_Strong (RNA) for intronic and silent variants, as well as missense/in-frame variants located outside a (potentially) clinically important functional domain. Missense variants located inside a (potentially) clinically important functional domain must meet BS3 to be eligible for BP7_Strong (RNA). See Specifications Figure1B and Appendix E for details.
352
+
353
+
354
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","General recommendation,Gene-specific"
355
+ BRCA1 (HGNC:1100),BP7,Supporting,"Silent variant inside a (potentially) clinically important functional domain, IF BP4 met.
356
+
357
+
358
+ Intronic variants located outside conserved donor or acceptor motif positions (at or beyond positions +7/-21) IF BP4 met.
359
+
360
+
361
+ See Specifications Figure1A and Appendix J for additional details.
362
+
363
+
364
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA1 RING aa 2-101; BRCA1 coiled-coil aa 1391-1424; BRCA1 BRCT repeats aa 1650-1857. See Specifications Figure1A and Appendix J for details.","Clarification,General recommendation"
VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA2Version1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,373 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ BRCA2 (HGNC:1101),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ BRCA2 (HGNC:1101),PVS1,Very Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
9
+
10
+
11
+ Well-established
12
+ in vitro
13
+ or
14
+ in vivo
15
+ functional studies supportive of a damaging effect
16
+ as measured by effect on mRNA transcript profile (mRNA assay only).
17
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
18
+ BRCA2 (HGNC:1101),PVS1,Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
19
+
20
+
21
+ Well-established
22
+ in vitro
23
+ or
24
+ in vivo
25
+ functional studies supportive of a damaging effect
26
+ as measured by effect on mRNA transcript profile (mRNA assay only).
27
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
28
+ BRCA2 (HGNC:1101),PVS1,Moderate,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
29
+
30
+
31
+ Well-established
32
+ in vitro
33
+ or
34
+ in vivo
35
+ functional studies supportive of a damaging effect
36
+ as measured by effect on mRNA transcript profile (mRNA assay only).
37
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
38
+ BRCA2 (HGNC:1101),PVS1,Supporting,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
39
+
40
+
41
+ Well-established
42
+ in vitro
43
+ or
44
+ in vivo
45
+ functional studies supportive of a damaging effect
46
+ as measured by effect on mRNA transcript profile (mRNA assay only).
47
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
48
+ BRCA2 (HGNC:1101),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
49
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
50
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
51
+ BRCA2 (HGNC:1101),PS1,Strong,"Apply
52
+ PS1
53
+ , for predicted
54
+ missense
55
+ substitutions, where a previously classified
56
+ pathogenic
57
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
58
+
59
+
60
+ Apply
61
+ PS1
62
+ , for exonic and intronic variants with same predicted impact on
63
+ splicing,
64
+ as a previously classified
65
+ pathogenic
66
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant. 
67
+
68
+
69
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
70
+ BRCA2 (HGNC:1101),PS1,Moderate,"Apply
71
+ PS1_Moderate
72
+ , for predicted
73
+ missense
74
+ substitutions, where a previously classified
75
+ likely pathogenic
76
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
77
+
78
+
79
+ Apply
80
+ PS1_Moderate
81
+ , for exonic and intronic variants with same predicted impact on
82
+ splicing
83
+ , as a previously classified
84
+ (likely) pathogenic
85
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
86
+
87
+
88
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
89
+ BRCA2 (HGNC:1101),PS1,Supporting,"Apply
90
+ PS1
91
+ , for exonic and intronic variants with same predicted impact on
92
+ splicing
93
+ , as a previously classified
94
+ (likely) pathogenic
95
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
96
+
97
+
98
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
99
+ BRCA2 (HGNC:1101),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
100
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
101
+ BRCA2 (HGNC:1101),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
102
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
103
+ BRCA2 (HGNC:1101),PS3,Strong,"Well-established in vitro or in vivo functional studies supportive of a damaging effect. Apply PS3 for assays measuring effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
104
+
105
+
106
+ Well-established
107
+ in vitro
108
+ or
109
+ in vivo
110
+ functional studies supportive of a damaging effect
111
+ as measured by effect on mRNA transcript profile (mRNA assay only).
112
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",General recommendation
113
+ BRCA2 (HGNC:1101),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
114
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
115
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
116
+ BRCA2 (HGNC:1101),PS4,Strong,The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls. Case-control studies; p-value ≤0.05 and OR ≥4 (lower confidence interval excludes 2.0). See Appendix F for details.,"Clarification,Gene-specific"
117
+ BRCA2 (HGNC:1101),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
118
+ BRCA2 (HGNC:1101),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
119
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
120
+ BRCA2 (HGNC:1101),PM2,Supporting,"Absent from controls in an outbred population, from gnomAD v2.1 (non-cancer, exome only subset) and gnomAD v3.1 (non-cancer). Region around the variant must have an average read depth ≥25. See Appendix G for details.",Gene-specific
121
+ BRCA2 (HGNC:1101),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
122
+ Note: This requires testing of parents (or offspring) to determine phase.",
123
+ BRCA2 (HGNC:1101),PM3,Strong,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene. Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
124
+
125
+
126
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
127
+ ≤5yr
128
+ .
129
+
130
+
131
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
132
+
133
+
134
+ See
135
+ Specifications Table 6
136
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
137
+
138
+
139
+ PM3_Strong = ≥4 points",Gene-specific
140
+ BRCA2 (HGNC:1101),PM3,Moderate,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene. Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
141
+
142
+
143
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
144
+ ≤5yr
145
+ .
146
+
147
+
148
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
149
+
150
+
151
+ See
152
+ Specifications Table 6
153
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
154
+
155
+
156
+ PM3 = 2 points",Gene-specific
157
+ BRCA2 (HGNC:1101),PM3,Supporting,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
158
+
159
+
160
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
161
+ ≤5yr
162
+ .
163
+
164
+
165
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
166
+
167
+
168
+ See
169
+ Specifications Table 6
170
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
171
+
172
+
173
+ PM3_Supporting = 1 point",Gene-specific
174
+ BRCA2 (HGNC:1101),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,NA
175
+ BRCA2 (HGNC:1101),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
176
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
177
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
178
+ BRCA2 (HGNC:1101),PM5,Strong,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
179
+ BRCA2 (HGNC:1101),PM5,Moderate,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
180
+ BRCA2 (HGNC:1101),PM5,Supporting,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
181
+ BRCA2 (HGNC:1101),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
182
+ BRCA2 (HGNC:1101),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
183
+ Note: May be used as stronger evidence with increasing segregation data.",
184
+ BRCA2 (HGNC:1101),PP1,Strong,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
185
+
186
+
187
+ Apply weight as per Bayes Score:
188
+
189
+
190
+ PP1_Strong – LR>18.7:1
191
+
192
+
193
+ PP1_Very Strong – LR>350:1",Gene-specific
194
+ BRCA2 (HGNC:1101),PP1,Moderate,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
195
+
196
+
197
+ Apply weight as per Bayes Score:
198
+
199
+
200
+ PP1_Moderate – LR>4.3:1",Gene-specific
201
+ BRCA2 (HGNC:1101),PP1,Supporting,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
202
+
203
+
204
+ Apply weight as per Bayes Score:
205
+
206
+
207
+ PP1 - LR >2.08:1",Gene-specific
208
+ BRCA2 (HGNC:1101),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
209
+ BRCA2 (HGNC:1101),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
210
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
211
+ BRCA2 (HGNC:1101),PP3,Supporting,"Apply PP3 for missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain and predicted impact via protein change (BayesDel no-AF score ≥0.30). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186.
212
+
213
+
214
+ Apply PP3 for predicted splicing (SpliceAI ≥0.2) for silent, missense/in-frame (irrespective of location in clinically important functional domain) and for intronic variants outside of donor and acceptor 1,2 sites.
215
+
216
+
217
+ See Specifications Figure1A and Appendix J for details.",Gene-specific
218
+ BRCA2 (HGNC:1101),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
219
+ BRCA2 (HGNC:1101),PP4,Strong,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
220
+
221
+
222
+ PP4_Strong – LR>18.7:1
223
+
224
+
225
+ PP4_Very Strong – LR>350:1
226
+
227
+
228
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
229
+
230
+
231
+ See Specifications Table7 and Appendix B for details.",Gene-specific
232
+ BRCA2 (HGNC:1101),PP4,Moderate,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
233
+
234
+
235
+ PP4_Moderate – LR>4.3:1
236
+
237
+
238
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
239
+
240
+
241
+ See Specifications Table7 and Appendix B for details.",Gene-specific
242
+ BRCA2 (HGNC:1101),PP4,Supporting,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
243
+
244
+
245
+ PP4 - LR >2.08:1 
246
+
247
+
248
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
249
+
250
+
251
+ See Specifications Table7 and Appendix B for details.",Gene-specific
252
+ BRCA2 (HGNC:1101),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
253
+ BRCA2 (HGNC:1101),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
254
+ BRCA2 (HGNC:1101),BA1,Stand Alone,"Filter allele frequency (FAF) is above 0.1% (FAF > 0.001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
255
+ BRCA2 (HGNC:1101),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
256
+ BRCA2 (HGNC:1101),BS1,Strong,"Filter allele frequency (FAF) is above 0.01% (FAF > 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
257
+ BRCA2 (HGNC:1101),BS1,Supporting,"Filter allele frequency (FAF) is above 0.002% (FAF > 0.00002) and less than or equal to 0.01% (FAF ≤ 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
258
+ BRCA2 (HGNC:1101),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
259
+ BRCA2 (HGNC:1101),BS2,Strong,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
260
+ Specifications Table 8
261
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
262
+
263
+
264
+ BS2 = ≥ 4 points",Gene-specific
265
+ BRCA2 (HGNC:1101),BS2,Moderate,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
266
+ Specifications Table 8
267
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
268
+
269
+
270
+ BS2_Moderate = 2 points",Gene-specific
271
+ BRCA2 (HGNC:1101),BS2,Supporting,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
272
+ Specifications Table 8
273
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
274
+
275
+
276
+ BS2_Supporting = 1 points",Gene-specific
277
+ BRCA2 (HGNC:1101),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
278
+ BRCA2 (HGNC:1101),BS3,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function. Assay measures effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
279
+
280
+
281
+ Well-established
282
+ in vitro
283
+ or
284
+ in vivo
285
+ functional studies supportive of no damaging effect
286
+ as measured by effect on mRNA transcript profile (mRNA assay only).
287
+ Apply as BP7 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
288
+ BRCA2 (HGNC:1101),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
289
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
290
+ BRCA2 (HGNC:1101),BS4,Strong,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
291
+
292
+
293
+ Apply weight as per Bayes Score:
294
+
295
+
296
+ BS4 - LR <0.05:1
297
+
298
+
299
+ BS4_VeryStrong – LR <0.00285:1",Gene-specific
300
+ BRCA2 (HGNC:1101),BS4,Moderate,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
301
+
302
+
303
+ Apply weight as per Bayes Score:
304
+
305
+
306
+ BS4_Moderate - LR <0.23:1",Gene-specific
307
+ BRCA2 (HGNC:1101),BS4,Supporting,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
308
+
309
+
310
+ Apply weight as per Bayes Score:
311
+
312
+
313
+ BS4_Supporting  - LR 0.23-0.48:1",Gene-specific
314
+ BRCA2 (HGNC:1101),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,
315
+ BRCA2 (HGNC:1101),BP1,Strong,"Apply BP1_Strong
316
+ for silent substitution, missense or in-frame insertion, deletion or delins variants outside a (potentially) clinically important functional domain AND no splicing predicted (SpliceAI ≤0.1). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","Gene-specific,Strength"
317
+ BRCA2 (HGNC:1101),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,NA
318
+ BRCA2 (HGNC:1101),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
319
+ BRCA2 (HGNC:1101),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
320
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
321
+ BRCA2 (HGNC:1101),BP4,Supporting,"Missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain, and no predicted impact via protein change or splicing (BayesDel no-AF score ≤ 0.18 AND SpliceAI ≤0.1).
322
+
323
+
324
+ Silent variant inside a (potentially) clinically important functional domain, if no predicted impact via splicing (SpliceAI ≤0.1).
325
+
326
+
327
+ Intronic variants outside of the native donor and acceptor splice sites (i.e. not +/- 1,2 positions) AND no predicted impact via splicing (SpliceAI ≤0.1).
328
+
329
+
330
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","Clarification,Gene-specific"
331
+ BRCA2 (HGNC:1101),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
332
+ BRCA2 (HGNC:1101),BP5,Strong,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
333
+
334
+
335
+ BP5_VeryStrong – LR <0.00285:1
336
+
337
+
338
+ BP5_Strong  - LR <0.05:1
339
+
340
+
341
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
342
+ BRCA2 (HGNC:1101),BP5,Moderate,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
343
+
344
+
345
+ BP5_Moderate - LR <0.23:1
346
+
347
+
348
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
349
+ BRCA2 (HGNC:1101),BP5,Supporting,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
350
+
351
+
352
+ BP5 - LR 0.23-0.48:1
353
+
354
+
355
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
356
+ BRCA2 (HGNC:1101),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
357
+ BRCA2 (HGNC:1101),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
358
+ BRCA2 (HGNC:1101),BP7,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function
359
+ as measured by effect on mRNA transcript profile – mRNA assay only.
360
+ Apply as BP7_Strong (RNA) for intronic and silent variants, as well as missense/in-frame variants located outside a (potentially) clinically important functional domain. See Specifications Figure1B and Appendix E for details.
361
+
362
+
363
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","General recommendation,Gene-specific"
364
+ BRCA2 (HGNC:1101),BP7,Supporting,"Silent variant inside a (potentially) clinically important functional domain, IF BP4 met.
365
+
366
+
367
+ Intronic variants located outside conserved donor or acceptor motif positions (at or beyond positions +7/-21) IF BP4 met.
368
+
369
+
370
+ See Specifications Figure1A and Appendix J for additional details.
371
+
372
+
373
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","Clarification,General recommendation"
VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA2Version1.1.0_version=1.1.0.csv ADDED
@@ -0,0 +1,373 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ BRCA2 (HGNC:1101),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ BRCA2 (HGNC:1101),PVS1,Very Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
9
+
10
+
11
+ Well-established
12
+ in vitro
13
+ or
14
+ in vivo
15
+ functional studies supportive of a damaging effect
16
+ as measured by effect on mRNA transcript profile (mRNA assay only).
17
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
18
+ BRCA2 (HGNC:1101),PVS1,Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
19
+
20
+
21
+ Well-established
22
+ in vitro
23
+ or
24
+ in vivo
25
+ functional studies supportive of a damaging effect
26
+ as measured by effect on mRNA transcript profile (mRNA assay only).
27
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
28
+ BRCA2 (HGNC:1101),PVS1,Moderate,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
29
+
30
+
31
+ Well-established
32
+ in vitro
33
+ or
34
+ in vivo
35
+ functional studies supportive of a damaging effect
36
+ as measured by effect on mRNA transcript profile (mRNA assay only).
37
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
38
+ BRCA2 (HGNC:1101),PVS1,Supporting,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
39
+
40
+
41
+ Well-established
42
+ in vitro
43
+ or
44
+ in vivo
45
+ functional studies supportive of a damaging effect
46
+ as measured by effect on mRNA transcript profile (mRNA assay only).
47
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
48
+ BRCA2 (HGNC:1101),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
49
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
50
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
51
+ BRCA2 (HGNC:1101),PS1,Strong,"Apply
52
+ PS1
53
+ , for predicted
54
+ missense
55
+ substitutions, where a previously classified
56
+ pathogenic
57
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
58
+
59
+
60
+ Apply
61
+ PS1
62
+ , for exonic and intronic variants with same predicted impact on
63
+ splicing,
64
+ as a previously classified
65
+ pathogenic
66
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant. 
67
+
68
+
69
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
70
+ BRCA2 (HGNC:1101),PS1,Moderate,"Apply
71
+ PS1_Moderate
72
+ , for predicted
73
+ missense
74
+ substitutions, where a previously classified
75
+ likely pathogenic
76
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
77
+
78
+
79
+ Apply
80
+ PS1_Moderate
81
+ , for exonic and intronic variants with same predicted impact on
82
+ splicing
83
+ , as a previously classified
84
+ (likely) pathogenic
85
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
86
+
87
+
88
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
89
+ BRCA2 (HGNC:1101),PS1,Supporting,"Apply
90
+ PS1
91
+ , for exonic and intronic variants with same predicted impact on
92
+ splicing
93
+ , as a previously classified
94
+ (likely) pathogenic
95
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
96
+
97
+
98
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
99
+ BRCA2 (HGNC:1101),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
100
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
101
+ BRCA2 (HGNC:1101),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
102
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
103
+ BRCA2 (HGNC:1101),PS3,Strong,"Well-established in vitro or in vivo functional studies supportive of a damaging effect. Apply PS3 for assays measuring effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
104
+
105
+
106
+ Well-established
107
+ in vitro
108
+ or
109
+ in vivo
110
+ functional studies supportive of a damaging effect
111
+ as measured by effect on mRNA transcript profile (mRNA assay only).
112
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",General recommendation
113
+ BRCA2 (HGNC:1101),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
114
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
115
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
116
+ BRCA2 (HGNC:1101),PS4,Strong,The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls. Case-control studies; p-value ≤0.05 and OR ≥4 (lower confidence interval excludes 2.0). See Appendix F for details.,"Clarification,Gene-specific"
117
+ BRCA2 (HGNC:1101),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
118
+ BRCA2 (HGNC:1101),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
119
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
120
+ BRCA2 (HGNC:1101),PM2,Supporting,"Absent from controls in an outbred population, from gnomAD v2.1 (non-cancer, exome only subset) and gnomAD v3.1 (non-cancer). Region around the variant must have an average read depth ≥25. See Appendix G for details.",Gene-specific
121
+ BRCA2 (HGNC:1101),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
122
+ Note: This requires testing of parents (or offspring) to determine phase.",
123
+ BRCA2 (HGNC:1101),PM3,Strong,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene. Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
124
+
125
+
126
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
127
+ ≤5yr
128
+ .
129
+
130
+
131
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
132
+
133
+
134
+ See
135
+ Specifications Table 6
136
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
137
+
138
+
139
+ PM3_Strong = ≥4 points",Gene-specific
140
+ BRCA2 (HGNC:1101),PM3,Moderate,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene. Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
141
+
142
+
143
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
144
+ ≤5yr
145
+ .
146
+
147
+
148
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
149
+
150
+
151
+ See
152
+ Specifications Table 6
153
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
154
+
155
+
156
+ PM3 = 2 points",Gene-specific
157
+ BRCA2 (HGNC:1101),PM3,Supporting,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
158
+
159
+
160
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
161
+ ≤5yr
162
+ .
163
+
164
+
165
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
166
+
167
+
168
+ See
169
+ Specifications Table 6
170
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
171
+
172
+
173
+ PM3_Supporting = 1 point",Gene-specific
174
+ BRCA2 (HGNC:1101),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,NA
175
+ BRCA2 (HGNC:1101),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
176
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
177
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
178
+ BRCA2 (HGNC:1101),PM5,Strong,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
179
+ BRCA2 (HGNC:1101),PM5,Moderate,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
180
+ BRCA2 (HGNC:1101),PM5,Supporting,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
181
+ BRCA2 (HGNC:1101),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
182
+ BRCA2 (HGNC:1101),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
183
+ Note: May be used as stronger evidence with increasing segregation data.",
184
+ BRCA2 (HGNC:1101),PP1,Strong,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
185
+
186
+
187
+ Apply weight as per Bayes Score:
188
+
189
+
190
+ PP1_Strong – LR>18.7:1
191
+
192
+
193
+ PP1_Very Strong – LR>350:1",Gene-specific
194
+ BRCA2 (HGNC:1101),PP1,Moderate,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
195
+
196
+
197
+ Apply weight as per Bayes Score:
198
+
199
+
200
+ PP1_Moderate – LR>4.3:1",Gene-specific
201
+ BRCA2 (HGNC:1101),PP1,Supporting,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
202
+
203
+
204
+ Apply weight as per Bayes Score:
205
+
206
+
207
+ PP1 - LR >2.08:1",Gene-specific
208
+ BRCA2 (HGNC:1101),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
209
+ BRCA2 (HGNC:1101),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
210
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
211
+ BRCA2 (HGNC:1101),PP3,Supporting,"Apply PP3 for missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain and predicted impact via protein change (BayesDel no-AF score ≥0.30). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186.
212
+
213
+
214
+ Apply PP3 for predicted splicing (SpliceAI ≥0.2) for silent, missense/in-frame (irrespective of location in clinically important functional domain) and for intronic variants outside of donor and acceptor 1,2 sites.
215
+
216
+
217
+ See Specifications Figure1A and Appendix J for details.",Gene-specific
218
+ BRCA2 (HGNC:1101),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
219
+ BRCA2 (HGNC:1101),PP4,Strong,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
220
+
221
+
222
+ PP4_Strong – LR>18.7:1
223
+
224
+
225
+ PP4_Very Strong – LR>350:1
226
+
227
+
228
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
229
+
230
+
231
+ See Specifications Table7 and Appendix B for details.",Gene-specific
232
+ BRCA2 (HGNC:1101),PP4,Moderate,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
233
+
234
+
235
+ PP4_Moderate – LR>4.3:1
236
+
237
+
238
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
239
+
240
+
241
+ See Specifications Table7 and Appendix B for details.",Gene-specific
242
+ BRCA2 (HGNC:1101),PP4,Supporting,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
243
+
244
+
245
+ PP4 - LR >2.08:1 
246
+
247
+
248
+ Combined LR 1.00-2.08 is not informative (PP4 not applicable).
249
+
250
+
251
+ See Specifications Table7 and Appendix B for details.",Gene-specific
252
+ BRCA2 (HGNC:1101),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
253
+ BRCA2 (HGNC:1101),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
254
+ BRCA2 (HGNC:1101),BA1,Stand Alone,"Filter allele frequency (FAF) is above 0.1% (FAF > 0.001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
255
+ BRCA2 (HGNC:1101),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
256
+ BRCA2 (HGNC:1101),BS1,Strong,"Filter allele frequency (FAF) is above 0.01% (FAF > 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
257
+ BRCA2 (HGNC:1101),BS1,Supporting,"Filter allele frequency (FAF) is above 0.002% (FAF > 0.00002) and less than or equal to 0.01% (FAF ≤ 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
258
+ BRCA2 (HGNC:1101),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
259
+ BRCA2 (HGNC:1101),BS2,Strong,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
260
+ Specifications Table 8
261
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
262
+
263
+
264
+ BS2 = ≥ 4 points",Gene-specific
265
+ BRCA2 (HGNC:1101),BS2,Moderate,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
266
+ Specifications Table 8
267
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
268
+
269
+
270
+ BS2_Moderate = 2 points",Gene-specific
271
+ BRCA2 (HGNC:1101),BS2,Supporting,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
272
+ Specifications Table 8
273
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
274
+
275
+
276
+ BS2_Supporting = 1 points",Gene-specific
277
+ BRCA2 (HGNC:1101),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
278
+ BRCA2 (HGNC:1101),BS3,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function. Assay measures effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
279
+
280
+
281
+ Well-established
282
+ in vitro
283
+ or
284
+ in vivo
285
+ functional studies supportive of no damaging effect
286
+ as measured by effect on mRNA transcript profile (mRNA assay only).
287
+ Apply as BP7 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
288
+ BRCA2 (HGNC:1101),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
289
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
290
+ BRCA2 (HGNC:1101),BS4,Strong,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
291
+
292
+
293
+ Apply weight as per Bayes Score:
294
+
295
+
296
+ BS4 - LR <0.05:1
297
+
298
+
299
+ BS4_VeryStrong – LR <0.00285:1",Gene-specific
300
+ BRCA2 (HGNC:1101),BS4,Moderate,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
301
+
302
+
303
+ Apply weight as per Bayes Score:
304
+
305
+
306
+ BS4_Moderate - LR <0.23:1",Gene-specific
307
+ BRCA2 (HGNC:1101),BS4,Supporting,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
308
+
309
+
310
+ Apply weight as per Bayes Score:
311
+
312
+
313
+ BS4_Supporting  - LR 0.23-0.48:1",Gene-specific
314
+ BRCA2 (HGNC:1101),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,
315
+ BRCA2 (HGNC:1101),BP1,Strong,"Apply BP1_Strong
316
+ for silent substitution, missense or in-frame insertion, deletion or delins variants outside a (potentially) clinically important functional domain AND no splicing predicted (SpliceAI ≤0.1). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","Gene-specific,Strength"
317
+ BRCA2 (HGNC:1101),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,NA
318
+ BRCA2 (HGNC:1101),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
319
+ BRCA2 (HGNC:1101),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
320
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
321
+ BRCA2 (HGNC:1101),BP4,Supporting,"Missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain, and no predicted impact via protein change or splicing (BayesDel no-AF score ≤ 0.18 AND SpliceAI ≤0.1).
322
+
323
+
324
+ Silent variant inside a (potentially) clinically important functional domain, if no predicted impact via splicing (SpliceAI ≤0.1).
325
+
326
+
327
+ Intronic variants outside of the native donor and acceptor splice sites (i.e. not +/- 1,2 positions) AND no predicted impact via splicing (SpliceAI ≤0.1).
328
+
329
+
330
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","Clarification,Gene-specific"
331
+ BRCA2 (HGNC:1101),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
332
+ BRCA2 (HGNC:1101),BP5,Strong,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
333
+
334
+
335
+ BP5_VeryStrong – LR <0.00285:1
336
+
337
+
338
+ BP5_Strong  - LR <0.05:1
339
+
340
+
341
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
342
+ BRCA2 (HGNC:1101),BP5,Moderate,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
343
+
344
+
345
+ BP5_Moderate - LR <0.23:1
346
+
347
+
348
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
349
+ BRCA2 (HGNC:1101),BP5,Supporting,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
350
+
351
+
352
+ BP5 - LR 0.23-0.48:1
353
+
354
+
355
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
356
+ BRCA2 (HGNC:1101),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
357
+ BRCA2 (HGNC:1101),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
358
+ BRCA2 (HGNC:1101),BP7,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function
359
+ as measured by effect on mRNA transcript profile – mRNA assay only.
360
+ Apply as BP7_Strong (RNA) for intronic and silent variants, as well as missense/in-frame variants located outside a (potentially) clinically important functional domain. See Specifications Figure1B and Appendix E for details.
361
+
362
+
363
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","General recommendation,Gene-specific"
364
+ BRCA2 (HGNC:1101),BP7,Supporting,"Silent variant inside a (potentially) clinically important functional domain, IF BP4 met.
365
+
366
+
367
+ Intronic variants located outside conserved donor or acceptor motif positions (at or beyond positions +7/-21) IF BP4 met.
368
+
369
+
370
+ See Specifications Figure1A and Appendix J for additional details.
371
+
372
+
373
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","Clarification,General recommendation"
VCI/parsing_csr_criteria/version_csv_individual/ClinGenENIGMABRCA1andBRCA2ExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforBRCA2Version1.2.0_version=1.2.0.csv ADDED
@@ -0,0 +1,364 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ BRCA2 (HGNC:1101),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ BRCA2 (HGNC:1101),PVS1,Very Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
9
+
10
+
11
+ Well-established
12
+ in vitro
13
+ or
14
+ in vivo
15
+ functional studies supportive of a damaging effect
16
+ as measured by effect on mRNA transcript profile (mRNA assay only).
17
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
18
+ BRCA2 (HGNC:1101),PVS1,Strong,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
19
+
20
+
21
+ Well-established
22
+ in vitro
23
+ or
24
+ in vivo
25
+ functional studies supportive of a damaging effect
26
+ as measured by effect on mRNA transcript profile (mRNA assay only).
27
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
28
+ BRCA2 (HGNC:1101),PVS1,Moderate,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
29
+
30
+
31
+ Well-established
32
+ in vitro
33
+ or
34
+ in vivo
35
+ functional studies supportive of a damaging effect
36
+ as measured by effect on mRNA transcript profile (mRNA assay only).
37
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
38
+ BRCA2 (HGNC:1101),PVS1,Supporting,"Null variant (nonsense, frameshift, splice site (donor/acceptor +/−1,2), initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease. Apply at appropriate strength according to PVS1 flowchart, which considers knowledge of clinically important functional domains. See Specifications Table 4 and Appendix D for details.
39
+
40
+
41
+ Well-established
42
+ in vitro
43
+ or
44
+ in vivo
45
+ functional studies supportive of a damaging effect
46
+ as measured by effect on mRNA transcript profile (mRNA assay only).
47
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
48
+ BRCA2 (HGNC:1101),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
49
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
50
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
51
+ BRCA2 (HGNC:1101),PS1,Strong,"Apply
52
+ PS1
53
+ , for predicted
54
+ missense
55
+ substitutions, where a previously classified
56
+ pathogenic
57
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
58
+
59
+
60
+ Apply
61
+ PS1
62
+ , for exonic and intronic variants with same predicted impact on
63
+ splicing,
64
+ as a previously classified
65
+ pathogenic
66
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant. 
67
+
68
+
69
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
70
+ BRCA2 (HGNC:1101),PS1,Moderate,"Apply
71
+ PS1_Moderate
72
+ , for predicted
73
+ missense
74
+ substitutions, where a previously classified
75
+ likely pathogenic
76
+ variant is considered to act via protein change (no confirmed or predicted effect on mRNA splicing (SpliceAI≤0.1)).
77
+
78
+
79
+ Apply
80
+ PS1_Moderate
81
+ , for exonic and intronic variants with same predicted impact on
82
+ splicing
83
+ , as a previously classified
84
+ (likely) pathogenic
85
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
86
+
87
+
88
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
89
+ BRCA2 (HGNC:1101),PS1,Supporting,"Apply
90
+ PS1
91
+ , for exonic and intronic variants with same predicted impact on
92
+ splicing
93
+ , as a previously classified
94
+ (likely) pathogenic
95
+ variant. Vary weight depending on relative positions, and confidence in classification of the reference variant.
96
+
97
+
98
+ See Specifications Table 5 and Appendix E, J and K for details.",General recommendation
99
+ BRCA2 (HGNC:1101),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
100
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",NA
101
+ BRCA2 (HGNC:1101),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
102
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
103
+ BRCA2 (HGNC:1101),PS3,Strong,"Well-established in vitro or in vivo functional studies supportive of a damaging effect. Apply PS3 for assays measuring effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
104
+
105
+
106
+ Well-established
107
+ in vitro
108
+ or
109
+ in vivo
110
+ functional studies supportive of a damaging effect
111
+ as measured by effect on mRNA transcript profile (mRNA assay only).
112
+ Apply as PVS1 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",General recommendation
113
+ BRCA2 (HGNC:1101),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
114
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
115
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
116
+ BRCA2 (HGNC:1101),PS4,Strong,The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls. Case-control studies; p-value ≤0.05 and OR ≥4 (lower confidence interval excludes 2.0). See Appendix F for details.,"Clarification,Gene-specific"
117
+ BRCA2 (HGNC:1101),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
118
+ BRCA2 (HGNC:1101),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
119
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
120
+ BRCA2 (HGNC:1101),PM2,Supporting,"Absent from controls in an outbred population, from gnomAD v2.1 (non-cancer, exome only subset) and gnomAD v3.1 (non-cancer). Region around the variant must have an average read depth ≥25. See Appendix G for details.",Gene-specific
121
+ BRCA2 (HGNC:1101),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
122
+ Note: This requires testing of parents (or offspring) to determine phase.",
123
+ BRCA2 (HGNC:1101),PM3,Strong,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene. Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
124
+
125
+
126
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
127
+ ≤5yr
128
+ .
129
+
130
+
131
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
132
+
133
+
134
+ See
135
+ Specifications Table 6
136
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
137
+
138
+
139
+ PM3_Strong = ≥4 points",Gene-specific
140
+ BRCA2 (HGNC:1101),PM3,Moderate,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene. Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
141
+
142
+
143
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
144
+ ≤5yr
145
+ .
146
+
147
+
148
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
149
+
150
+
151
+ See
152
+ Specifications Table 6
153
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
154
+
155
+
156
+ PM3 = 2 points",Gene-specific
157
+ BRCA2 (HGNC:1101),PM3,Supporting,"Apply for patient with phenotype consistent with BRCA1- or BRCA2-related Fanconi Anemia (FA), and co-occurrent variants in the same gene.Phenotype is considered consistent with BRCA1- or BRCA2-related FA if:
158
+
159
+
160
+ (i) Increased chromosome breakage (DEB, MMC, or spontaneous) and at least one clinical feature indicative of BRCA1/2-related FA, categorized under: physical features, pathology and laboratory findings, cancer diagnosis 
161
+ ≤5yr
162
+ .
163
+
164
+
165
+ (ii) Result unknown for chromosome breakage, and at least two clinical features indicative of BRCA1/2-related FA under at least two of the three categories: physical features, pathology and laboratory findings, cancer diagnosis ≤5yr.
166
+
167
+
168
+ See
169
+ Specifications Table 6
170
+ for approach to assign points per proband, and final PM3 code assignment based on the sum of PM3-related points. Also see Appendix H for additional details.
171
+
172
+
173
+ PM3_Supporting = 1 point",Gene-specific
174
+ BRCA2 (HGNC:1101),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,NA
175
+ BRCA2 (HGNC:1101),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
176
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
177
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
178
+ BRCA2 (HGNC:1101),PM5,Strong,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
179
+ BRCA2 (HGNC:1101),PM5,Moderate,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
180
+ BRCA2 (HGNC:1101),PM5,Supporting,Protein termination codon (PTC) variant in an exon where a different proven pathogenic PTC variant has been seen before. Use to justify additional weight for PTC variants annotated as PVS1. See Specifications Table 4 for PM5_PTC code strengths applicable per exon. See Appendix D for additional details.,Other
181
+ BRCA2 (HGNC:1101),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",NA
182
+ BRCA2 (HGNC:1101),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
183
+ Note: May be used as stronger evidence with increasing segregation data.",
184
+ BRCA2 (HGNC:1101),PP1,Strong,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
185
+
186
+
187
+ Apply weight as per Bayes Score:
188
+
189
+
190
+ PP1_Strong – LR ≥18.7:1
191
+
192
+
193
+ PP1_Very Strong – LR ≥350:1",Gene-specific
194
+ BRCA2 (HGNC:1101),PP1,Moderate,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
195
+
196
+
197
+ Apply weight as per Bayes Score:
198
+
199
+
200
+ PP1_Moderate – LR ≥4.3:1",Gene-specific
201
+ BRCA2 (HGNC:1101),PP1,Supporting,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
202
+
203
+
204
+ Apply weight as per Bayes Score:
205
+
206
+
207
+ PP1 - LR ≥2.08:1",Gene-specific
208
+ BRCA2 (HGNC:1101),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
209
+ BRCA2 (HGNC:1101),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
210
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
211
+ BRCA2 (HGNC:1101),PP3,Supporting,"Apply PP3 for missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain and predicted impact via protein change (BayesDel no-AF score ≥0.30). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186.
212
+
213
+
214
+ Apply PP3 for predicted splicing (SpliceAI ≥0.2) for silent, missense/in-frame (irrespective of location in clinically important functional domain) and for intronic variants outside of donor and acceptor 1,2 sites.
215
+
216
+
217
+ See Specifications Figure1A and Appendix J for details.",Gene-specific
218
+ BRCA2 (HGNC:1101),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,
219
+ BRCA2 (HGNC:1101),PP4,Strong,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
220
+
221
+
222
+ PP4_Strong – LR ≥18.7:1
223
+
224
+
225
+ PP4_Very Strong – LR ≥350:1
226
+
227
+
228
+ See Specifications Table7 and Appendix B for details.",Gene-specific
229
+ BRCA2 (HGNC:1101),PP4,Moderate,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
230
+
231
+
232
+ PP4_Moderate – LR ≥4.3:1
233
+
234
+
235
+ See Specifications Table7 and Appendix B for details.",Gene-specific
236
+ BRCA2 (HGNC:1101),PP4,Supporting,"Breast cancer is very common and has a high degree of genetic heterogeneity (caused by pathogenic variants in numerous genes). Use ONLY to capture combined LR towards pathogenicity, based on multifactorial likelihood clinical data.
237
+
238
+
239
+ PP4 - LR ≥2.08:1 
240
+
241
+
242
+ See Specifications Table7 and Appendix B for details.",Gene-specific
243
+ BRCA2 (HGNC:1101),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
244
+ BRCA2 (HGNC:1101),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
245
+ BRCA2 (HGNC:1101),BA1,Stand Alone,"Filter allele frequency (FAF) is above 0.1% (FAF > 0.001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
246
+ BRCA2 (HGNC:1101),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
247
+ BRCA2 (HGNC:1101),BS1,Strong,"Filter allele frequency (FAF) is above 0.01% (FAF > 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
248
+ BRCA2 (HGNC:1101),BS1,Supporting,"Filter allele frequency (FAF) is above 0.002% (FAF > 0.00002) and less than or equal to 0.01% (FAF ≤ 0.0001) in gnomAD v2.1 (non-cancer, exome only subset) and/or gnomAD v3.1 (non-cancer), non-founder population(s). See Appendix G for details.",Gene-specific
249
+ BRCA2 (HGNC:1101),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
250
+ BRCA2 (HGNC:1101),BS2,Strong,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
251
+ Specifications Table 8
252
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
253
+
254
+
255
+ BS2 = ≥ 4 points",Gene-specific
256
+ BRCA2 (HGNC:1101),BS2,Moderate,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
257
+ Specifications Table 8
258
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
259
+
260
+
261
+ BS2_Moderate = 2 points",Gene-specific
262
+ BRCA2 (HGNC:1101),BS2,Supporting,"Applied in absence of features of recessive disease, namely Fanconi Anemia phenotype. See
263
+ Specifications Table 8
264
+  for additional stipulations, and approach to assign points per proband, and final BS2 code assignment based on the sum of BS2-related points. See Appendix H for additional details.
265
+
266
+
267
+ BS2_Supporting = 1 points",Gene-specific
268
+ BRCA2 (HGNC:1101),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,
269
+ BRCA2 (HGNC:1101),BS3,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function. Assay measures effect via protein only OR mRNA and protein combined. See Specifications Table 9 for code recommendations from calibrated published assays. Also see Figure1C and Appendix E for details.
270
+
271
+
272
+ Well-established
273
+ in vitro
274
+ or
275
+ in vivo
276
+ functional studies supportive of no damaging effect
277
+ as measured by effect on mRNA transcript profile (mRNA assay only).
278
+ Apply as BP7 (RNA) at appropriate strength. See Specifications Figure1B and Appendix E for details.",Gene-specific
279
+ BRCA2 (HGNC:1101),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
280
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",
281
+ BRCA2 (HGNC:1101),BS4,Strong,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
282
+
283
+
284
+ Apply weight as per Bayes Score:
285
+
286
+
287
+ BS4 - LR ≤0.05:1
288
+
289
+
290
+ BS4_VeryStrong – LR ≤0.00285:1",Gene-specific
291
+ BRCA2 (HGNC:1101),BS4,Moderate,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
292
+
293
+
294
+ Apply weight as per Bayes Score:
295
+
296
+
297
+ BS4_Moderate - LR ≤0.23:1",Gene-specific
298
+ BRCA2 (HGNC:1101),BS4,Supporting,"Lack of segregation in affected members of a family, as measured by a quantitative co-segregation analysis method. See Appendix I for details.
299
+
300
+
301
+ Apply weight as per Bayes Score:
302
+
303
+
304
+ BS4_Supporting  - LR ≤0.48:1",Gene-specific
305
+ BRCA2 (HGNC:1101),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,
306
+ BRCA2 (HGNC:1101),BP1,Strong,"Apply BP1_Strong
307
+ for silent substitution, missense or in-frame insertion, deletion or delins variants outside a (potentially) clinically important functional domain AND no splicing predicted (SpliceAI ≤0.1). As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","Gene-specific,Strength"
308
+ BRCA2 (HGNC:1101),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,NA
309
+ BRCA2 (HGNC:1101),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,NA
310
+ BRCA2 (HGNC:1101),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
311
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
312
+ BRCA2 (HGNC:1101),BP4,Supporting,"Missense or in-frame insertion, deletion or delins variants inside a (potentially) clinically important functional domain, and no predicted impact via protein change or splicing (BayesDel no-AF score ≤ 0.18 AND SpliceAI ≤0.1).
313
+
314
+
315
+ Silent variant inside a (potentially) clinically important functional domain, if no predicted impact via splicing (SpliceAI ≤0.1).
316
+
317
+
318
+ Intronic variants outside of the native donor and acceptor splice sites (i.e. not +/- 1,2 positions) AND no predicted impact via splicing (SpliceAI ≤0.1).
319
+
320
+
321
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","Clarification,Gene-specific"
322
+ BRCA2 (HGNC:1101),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
323
+ BRCA2 (HGNC:1101),BP5,Strong,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
324
+
325
+
326
+ BP5_VeryStrong – LR ≤0.00285:1
327
+
328
+
329
+ BP5_Strong  - LR ≤0.05:1
330
+
331
+
332
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
333
+ BRCA2 (HGNC:1101),BP5,Moderate,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
334
+
335
+
336
+ BP5_Moderate - LR ≤0.23:1
337
+
338
+
339
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
340
+ BRCA2 (HGNC:1101),BP5,Supporting,"Use ONLY to capture combined LR against pathogenicity, based on multifactorial likelihood clinical data.
341
+
342
+
343
+ BP5 - LR ≤0.48:1
344
+
345
+
346
+ Not applicable for co-observation: cases with pathogenic variants in two (or more) different known breast–ovarian cancer risk genes have no specific phenotype.",Gene-specific
347
+ BRCA2 (HGNC:1101),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
348
+ BRCA2 (HGNC:1101),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
349
+ BRCA2 (HGNC:1101),BP7,Strong,"Well-established in vitro or in vivo functional studies shows no damaging effect on protein function
350
+ as measured by effect on mRNA transcript profile – mRNA assay only.
351
+ Apply as BP7_Strong (RNA) for intronic and silent variants, as well as missense/in-frame variants located outside a (potentially) clinically important functional domain. Missense variants located inside a (potentially) clinically important functional domain must meet BS3 to be eligible for BP7_Strong (RNA). See Specifications Figure1B and Appendix E for details.
352
+
353
+
354
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","General recommendation,Gene-specific"
355
+ BRCA2 (HGNC:1101),BP7,Supporting,"Silent variant inside a (potentially) clinically important functional domain, IF BP4 met.
356
+
357
+
358
+ Intronic variants located outside conserved donor or acceptor motif positions (at or beyond positions +7/-21) IF BP4 met.
359
+
360
+
361
+ See Specifications Figure1A and Appendix J for additional details.
362
+
363
+
364
+ As justified in the appendices, (potentially) clinically important functional domains are defined as: BRCA2 PALB2 binding domain aa 10-40; BRCA2 DNA binding aa 2481-3186. See Specifications Figure1A and Appendix J for details.","Clarification,General recommendation"
VCI/parsing_csr_criteria/version_csv_individual/ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1AVersion1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,521 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ SCN1A (HGNC:10585),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ SCN1A (HGNC:10585),PVS1,Very Strong,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”, using SCN1A-specific information.
9
+
10
+
11
+ Most terminal codon predicted to undergo nonsense-mediated decay (NMD) p.Thr1601. 
12
+
13
+
14
+ In-frame exons: 1, 7, 8, 12, 17, 25
15
+
16
+
17
+ Out-of-frame exons: 2-6, 9-11, 13-16, 18-24, 26
18
+
19
+
20
+ Truncated/altered protein is critical to protein funtion if the region falls within a Pathogenic Enriched Region, as defined in “PM1 Table”.
21
+
22
+
23
+ For splice region variants, this criterion should not be applied with PP3.","Disease-specific,General recommendation"
24
+ SCN1A (HGNC:10585),PVS1,Strong,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”, using SCN1A-specific information.
25
+
26
+
27
+ Most terminal codon predicted to undergo nonsense-mediated decay (NMD) p.Thr1601. 
28
+
29
+
30
+ In-frame exons: 1, 7, 8, 12, 17, 25
31
+
32
+
33
+ Out-of-frame exons: 2-6, 9-11, 13-16, 18-24, 26
34
+
35
+
36
+ Truncated/altered protein is critical to protein funtion if the region falls within a Pathogenic Enriched Region, as defined in “PM1 Table”.
37
+
38
+
39
+ For splice region variants, this criterion should not be applied with PP3.","Disease-specific,General recommendation"
40
+ SCN1A (HGNC:10585),PVS1,Moderate,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”, using SCN1A-specific information.
41
+
42
+
43
+ Most terminal codon predicted to undergo nonsense-mediated decay (NMD) p.Thr1601. 
44
+
45
+
46
+ In-frame exons: 1, 7, 8, 12, 17, 25
47
+
48
+
49
+ Out-of-frame exons: 2-6, 9-11, 13-16, 18-24, 26
50
+
51
+
52
+ Truncated/altered protein is critical to protein funtion if the region falls within a Pathogenic Enriched Region, as defined in “PM1 Table”.
53
+
54
+
55
+ For splice region variants, this criterion should not be applied with PP3.","Disease-specific,General recommendation"
56
+ SCN1A (HGNC:10585),PVS1,Supporting,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”, using SCN1A-specific information.
57
+
58
+
59
+ Most terminal codon predicted to undergo nonsense-mediated decay (NMD) p.Thr1601. 
60
+
61
+
62
+ In-frame exons: 1, 7, 8, 12, 17, 25
63
+
64
+
65
+ Out-of-frame exons: 2-6, 9-11, 13-16, 18-24, 26
66
+
67
+
68
+ Truncated/altered protein is critical to protein funtion if the region falls within a Pathogenic Enriched Region, as defined in “PM1 Table”.
69
+
70
+
71
+ For splice region variants, this criterion should not be applied with PP3.","Disease-specific,General recommendation"
72
+ SCN1A (HGNC:10585),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
73
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
74
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
75
+ SCN1A (HGNC:10585),PS1,Strong,"Same/identical amino acid change as previously reported (Caveat: beware of changes that impact splicing rather than at the amino acid/protein level).
76
+
77
+
78
+
79
+
80
+ Same amino acid change as a previously established
81
+ Pathogenic
82
+ variant regardless of nucleotide change. Example: Val->Leu caused by either G>C or G>T in the same codon.
83
+
84
+
85
+ >1
86
+ Identical amino acid change in paralogous gene previously established as
87
+ Pathogenic or Likely Pathogenic
88
+ , including NDD genes with equivalent constraint scores (SCN1A, SCN2A, SCN3A, SCN8A). See Paralogous Gene Table for corresponding amino acid positions.
89
+
90
+
91
+
92
+
93
+ Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023). 
94
+
95
+
96
+
97
+
98
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").",Gene-specific
99
+ SCN1A (HGNC:10585),PS1,Moderate,"Same/identical amino acid change as previously reported (Caveat: beware of changes that impact splicing rather than at the amino acid/protein level).
100
+
101
+
102
+
103
+
104
+ Same amino acid change as a previously established
105
+ Likely Pathogenic
106
+ variant regardless of nucleotide change. Example: Val->Leu caused by either G>C or G>T in the same codon.
107
+
108
+
109
+
110
+
111
+ Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023).
112
+
113
+
114
+
115
+
116
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").","Gene-specific,Strength"
117
+ SCN1A (HGNC:10585),PS1,Supporting,"Same/identical amino acid change as previously reported (Caveat: beware of changes that impact splicing rather than at the amino acid/protein level).
118
+
119
+
120
+
121
+
122
+ A single identical amino acid change in a paralogous gene previously established as
123
+ Pathogenic or Likely Pathogenic
124
+ , including NDD genes with equivalent constraint scores (SCN1A, SCN2A, SCN3A, SCN8A). See Paralogous Gene Table for corresponding amino acid positions.
125
+
126
+
127
+
128
+
129
+ Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023).
130
+
131
+
132
+
133
+
134
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").",Gene-specific
135
+ SCN1A (HGNC:10585),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
136
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
137
+ SCN1A (HGNC:10585),PS2,Very Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
138
+ 4 points
139
+ will arrive at
140
+ Very Strong
141
+
142
+
143
+
144
+ Dravet*: 2 points
145
+
146
+
147
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
148
+
149
+
150
+ Developmental and Epileptic Encephalopathy: 1 point
151
+
152
+
153
+ Hemiplegic migraine: 0.5 points
154
+
155
+
156
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
157
+
158
+
159
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
160
+ SCN1A (HGNC:10585),PS2,Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
161
+ 2 points
162
+ will arrive at
163
+ Strong
164
+
165
+
166
+
167
+ Dravet*: 2 points
168
+
169
+
170
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
171
+
172
+
173
+ Developmental and Epileptic Encephalopathy: 1 point
174
+
175
+
176
+ Hemiplegic migraine: 0.5 points
177
+
178
+
179
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
180
+
181
+
182
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
183
+ SCN1A (HGNC:10585),PS2,Moderate,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
184
+ 1 point
185
+ will arrive at
186
+ Moderate
187
+
188
+
189
+
190
+ Dravet*: 2 points
191
+
192
+
193
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
194
+
195
+
196
+ Developmental and Epileptic Encephalopathy: 1 point
197
+
198
+
199
+ Hemiplegic migraine: 0.5 points
200
+
201
+
202
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
203
+
204
+
205
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
206
+ SCN1A (HGNC:10585),PS2,Supporting,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
207
+ 0.5 points
208
+ will arrive at
209
+ Supporting
210
+
211
+
212
+
213
+ Dravet*: 2 points
214
+
215
+
216
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
217
+
218
+
219
+ Developmental and Epileptic Encephalopathy: 1 point
220
+
221
+
222
+ Hemiplegic migraine: 0.5 points
223
+
224
+
225
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
226
+
227
+
228
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
229
+ SCN1A (HGNC:10585),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
230
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
231
+ SCN1A (HGNC:10585),PS3,Strong,"In patch clamping experiments: Peak current as defined by FENICS ontology (
232
+ https://bioportal.bioontology.org/ontologies/FENICS
233
+ ) is less than or equal to 72.7% of wildtype.
234
+
235
+
236
+ In patch clamping experiments: Persistent current as defined by FENICS ontology (
237
+ https://bioportal.bioontology.org/ontologies/FENICS
238
+ ) is greater than or equal to 135% of wildtype.
239
+
240
+
241
+ In patch clamping experiments: Voltage dependence of activation as defined by FENICS ontology (
242
+ https://bioportal.bioontology.org/ontologies/FENICS
243
+ ) is shifted by at least 2.2 mV (absolute value).
244
+
245
+
246
+ In patch clamping experiments: Voltage dependence of inactivation as defined by FENICS ontology (
247
+ https://bioportal.bioontology.org/ontologies/FENICS
248
+ ) is shifted by at least 4.1 mV (absolute value).
249
+
250
+
251
+ Mouse knock-in model displays spontaneous seizures.","Disease-specific,Gene-specific"
252
+ SCN1A (HGNC:10585),PS3,Moderate,"In patch clamping experiments: Peak current as defined by FENICS ontology (
253
+ https://bioportal.bioontology.org/ontologies/FENICS
254
+ ) is less than or equal to 80.6% of wildtype.
255
+
256
+
257
+ In patch clamping experiments: Persistent current as defined by FENICS ontology (
258
+ https://bioportal.bioontology.org/ontologies/FENICS
259
+ ) is greater than or equal to 125% of wildtype.
260
+
261
+
262
+ In patch clamping experiments: Voltage dependence of activation as defined by FENICS ontology (
263
+ https://bioportal.bioontology.org/ontologies/FENICS
264
+ ) is shifted by at least 2.1 mV (absolute value).
265
+
266
+
267
+ In patch clamping experiments: Voltage dependence of inactivation as defined by FENICS ontology (
268
+ https://bioportal.bioontology.org/ontologies/FENICS
269
+ ) is shifted by at least 3.0 mV (absolute value).
270
+
271
+
272
+ Mouse knock-in model displays induced seizures
273
+
274
+
275
+ Zebrafish knock-in model displays spontaneous seizures, evidenced by hyperexcitability through electrophysiology or calcium imaging-based studies","Gene-specific,Strength"
276
+ SCN1A (HGNC:10585),PS3,Supporting,"Zebrafish knock-in model displays induced seizures, evidenced by hyperexcitability through electrophysiology or calcium imaging-based studies","Gene-specific,Strength"
277
+ SCN1A (HGNC:10585),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
278
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
279
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
280
+ SCN1A (HGNC:10585),PS4,Very Strong,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
281
+ 16+ points
282
+ will arrive at
283
+ Very Strong
284
+
285
+
286
+
287
+ Dravet*: 2 points
288
+
289
+
290
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
291
+
292
+
293
+ Developmental and Epileptic Encephalopathy: 1 point
294
+
295
+
296
+ Hemiplegic migraine: 0.5 points
297
+
298
+
299
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
300
+ SCN1A (HGNC:10585),PS4,Strong,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
301
+ 4-15.5 points
302
+ will arrive at
303
+ Strong
304
+
305
+
306
+
307
+ Dravet*: 2 points
308
+
309
+
310
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
311
+
312
+
313
+ Developmental and Epileptic Encephalopathy: 1 point
314
+
315
+
316
+ Hemiplegic migraine: 0.5 points
317
+
318
+
319
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific"
320
+ SCN1A (HGNC:10585),PS4,Moderate,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
321
+ 2-3.5 points
322
+ will arrive at
323
+ Moderate
324
+
325
+
326
+
327
+ Dravet*: 2 points
328
+
329
+
330
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
331
+
332
+
333
+ Developmental and Epileptic Encephalopathy: 1 point
334
+
335
+
336
+ Hemiplegic migraine: 0.5 points
337
+
338
+
339
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
340
+ SCN1A (HGNC:10585),PS4,Supporting,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
341
+ 1-1.5 points
342
+ will arrive at
343
+ Supporting
344
+
345
+
346
+
347
+ Dravet*: 2 points
348
+
349
+
350
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
351
+
352
+
353
+ Developmental and Epileptic Encephalopathy: 1 point
354
+
355
+
356
+ Hemiplegic migraine: 0.5 points
357
+
358
+
359
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
360
+ SCN1A (HGNC:10585),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
361
+ SCN1A (HGNC:10585),PM1,Moderate,"Variant is located within a Pathogenic Enriched Region. See specific amino acid residues noted in the attached “PM1 Table"".",Gene-specific
362
+ SCN1A (HGNC:10585),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
363
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
364
+ SCN1A (HGNC:10585),PM2,Supporting,"One or fewer alleles, if a minimum of 10,000 alleles assessed in population databases, such as the Genome Aggregation Database (gnomAD). Caveat: Population data for indels may be poorly called by next generation sequencing.","General recommendation,Other"
365
+ SCN1A (HGNC:10585),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
366
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
367
+ SCN1A (HGNC:10585),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
368
+ SCN1A (HGNC:10585),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
369
+ SCN1A (HGNC:10585),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
370
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
371
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
372
+ SCN1A (HGNC:10585),PM5,Strong,Greater than or equal to 2 known pathogenic variants at same site as novel change (within the same gene).,"Disease-specific,Gene-specific,Strength"
373
+ SCN1A (HGNC:10585),PM5,Moderate,Novel missense change at an amino acid residue in the same gene where a different missense variant was determined to be Pathogenic.,General recommendation
374
+ SCN1A (HGNC:10585),PM5,Supporting,"Novel missense change at an amino acid residue where a different missense change determined to be
375
+ Likely Pathogenic
376
+ has been seen before. Example: Arg156His is pathogenic; now you observe Arg156Cys. 
377
+
378
+
379
+ >1 Non-Identical aa change in paralogous gene(s) where a different missense change determined to be
380
+ Pathogenic
381
+ or
382
+ Likely Pathogenic
383
+ , including NDD genes with equivalent constraint scores (SCN1A, SCN2A, SCN3A, SCN8A)
384
+
385
+
386
+
387
+
388
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.","Disease-specific,Gene-specific,Strength"
389
+ SCN1A (HGNC:10585),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
390
+ SCN1A (HGNC:10585),PM6,Strong,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
391
+ 2 points
392
+ will arrive at
393
+ Strong
394
+ . Total of
395
+ 4 points
396
+ will arrive at
397
+ Very Strong
398
+
399
+
400
+
401
+ Dravet*: 1 points
402
+
403
+
404
+ Genetic Epilepsy with Febrile Seizures Plus: 0.5 points
405
+
406
+
407
+ Developmental and Epileptic Encephalopathy: 0.5 points
408
+
409
+
410
+ Hemiplegic migraine: 0.25 points
411
+
412
+
413
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
414
+ SCN1A (HGNC:10585),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
415
+ 1 point
416
+ will arrive at
417
+ Moderate
418
+
419
+
420
+
421
+ Dravet*: 1 points
422
+
423
+
424
+ Genetic Epilepsy with Febrile Seizures Plus: 0.5 points
425
+
426
+
427
+ Developmental and Epileptic Encephalopathy: 0.5 points
428
+
429
+
430
+ Hemiplegic migraine: 0.25 points
431
+
432
+
433
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
434
+ SCN1A (HGNC:10585),PM6,Supporting,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
435
+ 0.5 points
436
+ will arrive at
437
+ Supporting
438
+
439
+
440
+
441
+ Dravet*: 1 points
442
+
443
+
444
+ Genetic Epilepsy with Febrile Seizures Plus: 0.5 points
445
+
446
+
447
+ Developmental and Epileptic Encephalopathy: 0.5 points
448
+
449
+
450
+ Hemiplegic migraine: 0.25 points
451
+
452
+
453
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
454
+ SCN1A (HGNC:10585),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
455
+ Note: May be used as stronger evidence with increasing segregation data.",
456
+ SCN1A (HGNC:10585),PP1,Strong,"Co-segregation with disease in multiple affected family members.
457
+
458
+
459
+ >=7 independent meioses",Strength
460
+ SCN1A (HGNC:10585),PP1,Moderate,"Co-segregation with disease in multiple affected family members.
461
+
462
+
463
+ 5-6 independent meioses",Strength
464
+ SCN1A (HGNC:10585),PP1,Supporting,"Co-segregation with disease in multiple affected family members.
465
+
466
+
467
+ 3-4 independent meioses",Strength
468
+ SCN1A (HGNC:10585),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
469
+ SCN1A (HGNC:10585),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
470
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
471
+ SCN1A (HGNC:10585),PP3,Moderate,"Follow ClinGen’s recommendations (
472
+ PMID: 36413997
473
+ ) using REVEL as the computational tool, with the following stipulations:
474
+
475
+
476
+
477
+
478
+ Strength should be capped at Moderate, and 
479
+
480
+
481
+ limit the combination of PP3 and PM1 to reach no higher than strong","General recommendation,Strength"
482
+ SCN1A (HGNC:10585),PP3,Supporting,"Follow ClinGen’s recommendations (
483
+ PMID: 36413997
484
+ ) using REVEL as the computational tool, with the following stipulations: with the following stipulations:
485
+
486
+
487
+
488
+
489
+ Strength should be capped at Moderate, and 
490
+
491
+
492
+ limit the combination of PP3 and PM1 to reach no higher than strong","General recommendation,Strength"
493
+ SCN1A (HGNC:10585),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
494
+ SCN1A (HGNC:10585),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
495
+ SCN1A (HGNC:10585),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
496
+ SCN1A (HGNC:10585),BA1,Stand Alone,"Allele frequency is above 0.02% in GnomAD or other large population database, must be greater than or equal to 5 alleles if a minimum of 10,000 alleles was assessed.",Disease-specific
497
+ SCN1A (HGNC:10585),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
498
+ SCN1A (HGNC:10585),BS1,Strong,"Allele frequency is above 0.0004% in GnomAD or other large population database, must be greater than or equal to 5 alleles if a minimum of 10,000 alleles was assessed.",Disease-specific
499
+ SCN1A (HGNC:10585),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
500
+ SCN1A (HGNC:10585),BS2,Strong,Observed in a healthy adult individual.,No change
501
+ SCN1A (HGNC:10585),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,NA
502
+ SCN1A (HGNC:10585),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
503
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
504
+ SCN1A (HGNC:10585),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
505
+ SCN1A (HGNC:10585),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
506
+ SCN1A (HGNC:10585),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,No change
507
+ SCN1A (HGNC:10585),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,
508
+ SCN1A (HGNC:10585),BP3,Supporting,In frame-deletions/insertions in a repetitive region without a known function.,No change
509
+ SCN1A (HGNC:10585),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
510
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
511
+ SCN1A (HGNC:10585),BP4,Moderate,"Follow ClinGen’s recommendations (
512
+ PMID: 36413997
513
+ ) using REVEL as the computational tool.",General recommendation
514
+ SCN1A (HGNC:10585),BP4,Supporting,"Follow ClinGen’s recommendations (
515
+ PMID: 36413997
516
+ ) using REVEL as the computational tool.",General recommendation
517
+ SCN1A (HGNC:10585),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
518
+ SCN1A (HGNC:10585),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease.,No change
519
+ SCN1A (HGNC:10585),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
520
+ SCN1A (HGNC:10585),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
521
+ SCN1A (HGNC:10585),BP7,Supporting,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,No change
VCI/parsing_csr_criteria/version_csv_individual/ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1AVersion2.0.0_version=2.0.0.csv ADDED
@@ -0,0 +1,461 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ SCN1A (HGNC:10585),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ SCN1A (HGNC:10585),PVS1,Very Strong,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”.","Disease-specific,General recommendation"
9
+ SCN1A (HGNC:10585),PVS1,Strong,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”.","Disease-specific,General recommendation"
10
+ SCN1A (HGNC:10585),PVS1,Moderate,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”.","Disease-specific,General recommendation"
11
+ SCN1A (HGNC:10585),PVS1,Supporting,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”.","Disease-specific,General recommendation"
12
+ SCN1A (HGNC:10585),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
13
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
14
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
15
+ SCN1A (HGNC:10585),PS1,Strong,"Same/identical amino acid change as previously reported (Caveat: beware of changes that impact splicing rather than at the amino acid/protein level).
16
+
17
+
18
+
19
+
20
+ Same amino acid change as a previously established
21
+ Pathogenic
22
+ variant regardless of nucleotide change. Example: Val->Leu caused by either G>C or G>T in the same codon.
23
+
24
+
25
+ >1
26
+ Identical amino acid change in paralogous gene previously established as
27
+ Pathogenic or Likely Pathogenic
28
+ , including NDD genes with equivalent constraint scores (SCN1A, SCN2A, SCN3A, SCN8A). See Paralogous Gene Table for corresponding amino acid positions.
29
+
30
+
31
+
32
+
33
+ Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023). 
34
+
35
+
36
+
37
+
38
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").",Gene-specific
39
+ SCN1A (HGNC:10585),PS1,Moderate,"Same/identical amino acid change as previously reported (Caveat: beware of changes that impact splicing rather than at the amino acid/protein level).
40
+
41
+
42
+
43
+
44
+ Same amino acid change as a previously established
45
+ Likely Pathogenic
46
+ variant regardless of nucleotide change. Example: Val->Leu caused by either G>C or G>T in the same codon.
47
+
48
+
49
+
50
+
51
+ Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023).
52
+
53
+
54
+
55
+
56
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").","Gene-specific,Strength"
57
+ SCN1A (HGNC:10585),PS1,Supporting,"Same/identical amino acid change as previously reported (Caveat: beware of changes that impact splicing rather than at the amino acid/protein level).
58
+
59
+
60
+
61
+
62
+ A single identical amino acid change in a paralogous gene previously established as
63
+ Pathogenic or Likely Pathogenic
64
+ , including NDD genes with equivalent constraint scores (SCN1A, SCN2A, SCN3A, SCN8A). See Paralogous Gene Table for corresponding amino acid positions.
65
+
66
+
67
+
68
+
69
+ Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023).
70
+
71
+
72
+
73
+
74
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").",Gene-specific
75
+ SCN1A (HGNC:10585),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
76
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
77
+ SCN1A (HGNC:10585),PS2,Very Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
78
+ 4 points
79
+ will arrive at
80
+ Very Strong
81
+
82
+
83
+
84
+ Dravet*: 2 points
85
+
86
+
87
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
88
+
89
+
90
+ Developmental and Epileptic Encephalopathy: 1 point
91
+
92
+
93
+ Hemiplegic migraine: 0.5 points
94
+
95
+
96
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
97
+
98
+
99
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
100
+ SCN1A (HGNC:10585),PS2,Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
101
+ 2 points
102
+ will arrive at
103
+ Strong
104
+
105
+
106
+
107
+ Dravet*: 2 points
108
+
109
+
110
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
111
+
112
+
113
+ Developmental and Epileptic Encephalopathy: 1 point
114
+
115
+
116
+ Hemiplegic migraine: 0.5 points
117
+
118
+
119
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
120
+
121
+
122
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
123
+ SCN1A (HGNC:10585),PS2,Moderate,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
124
+ 1 point
125
+ will arrive at
126
+ Moderate
127
+
128
+
129
+
130
+ Dravet*: 2 points
131
+
132
+
133
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
134
+
135
+
136
+ Developmental and Epileptic Encephalopathy: 1 point
137
+
138
+
139
+ Hemiplegic migraine: 0.5 points
140
+
141
+
142
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
143
+
144
+
145
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
146
+ SCN1A (HGNC:10585),PS2,Supporting,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
147
+ 0.5 points
148
+ will arrive at
149
+ Supporting
150
+
151
+
152
+
153
+ Dravet*: 2 points
154
+
155
+
156
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
157
+
158
+
159
+ Developmental and Epileptic Encephalopathy: 1 point
160
+
161
+
162
+ Hemiplegic migraine: 0.5 points
163
+
164
+
165
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
166
+
167
+
168
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
169
+ SCN1A (HGNC:10585),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
170
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
171
+ SCN1A (HGNC:10585),PS3,Strong,"In patch clamping experiments: Peak current as defined by FENICS ontology (
172
+ https://bioportal.bioontology.org/ontologies/FENICS
173
+ ) is less than or equal to 72.7% of wildtype.
174
+
175
+
176
+ In patch clamping experiments: Persistent current as defined by FENICS ontology (
177
+ https://bioportal.bioontology.org/ontologies/FENICS
178
+ ) is greater than or equal to 135% of wildtype.
179
+
180
+
181
+ In patch clamping experiments: Voltage dependence of activation as defined by FENICS ontology (
182
+ https://bioportal.bioontology.org/ontologies/FENICS
183
+ ) is shifted by at least 2.2 mV (absolute value).
184
+
185
+
186
+ In patch clamping experiments: Voltage dependence of inactivation as defined by FENICS ontology (
187
+ https://bioportal.bioontology.org/ontologies/FENICS
188
+ ) is shifted by at least 4.1 mV (absolute value).
189
+
190
+
191
+ Mouse knock-in model displays spontaneous seizures.","Disease-specific,Gene-specific"
192
+ SCN1A (HGNC:10585),PS3,Moderate,"In patch clamping experiments: Peak current as defined by FENICS ontology (
193
+ https://bioportal.bioontology.org/ontologies/FENICS
194
+ ) is less than or equal to 80.6% of wildtype.
195
+
196
+
197
+ In patch clamping experiments: Persistent current as defined by FENICS ontology (
198
+ https://bioportal.bioontology.org/ontologies/FENICS
199
+ ) is greater than or equal to 125% of wildtype.
200
+
201
+
202
+ In patch clamping experiments: Voltage dependence of activation as defined by FENICS ontology (
203
+ https://bioportal.bioontology.org/ontologies/FENICS
204
+ ) is shifted by at least 2.1 mV (absolute value).
205
+
206
+
207
+ In patch clamping experiments: Voltage dependence of inactivation as defined by FENICS ontology (
208
+ https://bioportal.bioontology.org/ontologies/FENICS
209
+ ) is shifted by at least 3.0 mV (absolute value).
210
+
211
+
212
+ Mouse knock-in model displays induced seizures
213
+
214
+
215
+ Zebrafish knock-in model displays spontaneous seizures, evidenced by hyperexcitability through electrophysiology or calcium imaging-based studies","Gene-specific,Strength"
216
+ SCN1A (HGNC:10585),PS3,Supporting,"Zebrafish knock-in model displays induced seizures, evidenced by hyperexcitability through electrophysiology or calcium imaging-based studies","Gene-specific,Strength"
217
+ SCN1A (HGNC:10585),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
218
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
219
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
220
+ SCN1A (HGNC:10585),PS4,Very Strong,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
221
+ 16+ points
222
+ will arrive at
223
+ Very Strong
224
+
225
+
226
+
227
+ Dravet*: 2 points
228
+
229
+
230
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
231
+
232
+
233
+ Developmental and Epileptic Encephalopathy: 1 point
234
+
235
+
236
+ Hemiplegic migraine: 0.5 points
237
+
238
+
239
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
240
+ SCN1A (HGNC:10585),PS4,Strong,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
241
+ 4-15.5 points
242
+ will arrive at
243
+ Strong
244
+
245
+
246
+
247
+ Dravet*: 2 points
248
+
249
+
250
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
251
+
252
+
253
+ Developmental and Epileptic Encephalopathy: 1 point
254
+
255
+
256
+ Hemiplegic migraine: 0.5 points
257
+
258
+
259
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific"
260
+ SCN1A (HGNC:10585),PS4,Moderate,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
261
+ 2-3.5 points
262
+ will arrive at
263
+ Moderate
264
+
265
+
266
+
267
+ Dravet*: 2 points
268
+
269
+
270
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
271
+
272
+
273
+ Developmental and Epileptic Encephalopathy: 1 point
274
+
275
+
276
+ Hemiplegic migraine: 0.5 points
277
+
278
+
279
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
280
+ SCN1A (HGNC:10585),PS4,Supporting,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
281
+ 1-1.5 points
282
+ will arrive at
283
+ Supporting
284
+
285
+
286
+
287
+ Dravet*: 2 points
288
+
289
+
290
+ Genetic Epilepsy with Febrile Seizures Plus: 1 point
291
+
292
+
293
+ Developmental and Epileptic Encephalopathy: 1 point
294
+
295
+
296
+ Hemiplegic migraine: 0.5 points
297
+
298
+
299
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
300
+ SCN1A (HGNC:10585),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,
301
+ SCN1A (HGNC:10585),PM1,Moderate,"Variant is located within a Pathogenic Enriched Region. See specific amino acid residues noted in the attached “PM1 Table"".",Gene-specific
302
+ SCN1A (HGNC:10585),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
303
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
304
+ SCN1A (HGNC:10585),PM2,Supporting,"One or fewer alleles, if a minimum of 10,000 alleles assessed in population databases, such as the Genome Aggregation Database (gnomAD). Caveat: Population data for indels may be poorly called by next generation sequencing.","General recommendation,Other"
305
+ SCN1A (HGNC:10585),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
306
+ Note: This requires testing of parents (or offspring) to determine phase.",NA
307
+ SCN1A (HGNC:10585),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
308
+ SCN1A (HGNC:10585),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
309
+ SCN1A (HGNC:10585),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
310
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
311
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
312
+ SCN1A (HGNC:10585),PM5,Strong,Greater than or equal to 2 known pathogenic variants at same site as novel change (within the same gene).,"Disease-specific,Gene-specific,Strength"
313
+ SCN1A (HGNC:10585),PM5,Moderate,Novel missense change at an amino acid residue in the same gene where a different missense variant was determined to be Pathogenic.,General recommendation
314
+ SCN1A (HGNC:10585),PM5,Supporting,"Novel missense change at an amino acid residue where a different missense change determined to be
315
+ Likely Pathogenic
316
+ has been seen before. Example: Arg156His is pathogenic; now you observe Arg156Cys. 
317
+
318
+
319
+ >1 Non-Identical aa change in paralogous gene(s) where a different missense change determined to be
320
+ Pathogenic
321
+ or
322
+ Likely Pathogenic
323
+ , including NDD genes with equivalent constraint scores (SCN1A, SCN2A, SCN3A, SCN8A)
324
+
325
+
326
+
327
+
328
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.","Disease-specific,Gene-specific,Strength"
329
+ SCN1A (HGNC:10585),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
330
+ SCN1A (HGNC:10585),PM6,Strong,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
331
+ 2 points
332
+ will arrive at
333
+ Strong
334
+ . Total of
335
+ 4 points
336
+ will arrive at
337
+ Very Strong
338
+
339
+
340
+
341
+ Dravet*: 1 points
342
+
343
+
344
+ Genetic Epilepsy with Febrile Seizures Plus: 0.5 points
345
+
346
+
347
+ Developmental and Epileptic Encephalopathy: 0.5 points
348
+
349
+
350
+ Hemiplegic migraine: 0.25 points
351
+
352
+
353
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
354
+ SCN1A (HGNC:10585),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
355
+ 1 point
356
+ will arrive at
357
+ Moderate
358
+
359
+
360
+
361
+ Dravet*: 1 points
362
+
363
+
364
+ Genetic Epilepsy with Febrile Seizures Plus: 0.5 points
365
+
366
+
367
+ Developmental and Epileptic Encephalopathy: 0.5 points
368
+
369
+
370
+ Hemiplegic migraine: 0.25 points
371
+
372
+
373
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
374
+ SCN1A (HGNC:10585),PM6,Supporting,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
375
+ 0.5 points
376
+ will arrive at
377
+ Supporting
378
+
379
+
380
+
381
+ Dravet*: 1 points
382
+
383
+
384
+ Genetic Epilepsy with Febrile Seizures Plus: 0.5 points
385
+
386
+
387
+ Developmental and Epileptic Encephalopathy: 0.5 points
388
+
389
+
390
+ Hemiplegic migraine: 0.25 points
391
+
392
+
393
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
394
+ SCN1A (HGNC:10585),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
395
+ Note: May be used as stronger evidence with increasing segregation data.",
396
+ SCN1A (HGNC:10585),PP1,Strong,"Co-segregation with disease in multiple affected family members.
397
+
398
+
399
+ >=7 independent meioses",Strength
400
+ SCN1A (HGNC:10585),PP1,Moderate,"Co-segregation with disease in multiple affected family members.
401
+
402
+
403
+ 5-6 independent meioses",Strength
404
+ SCN1A (HGNC:10585),PP1,Supporting,"Co-segregation with disease in multiple affected family members.
405
+
406
+
407
+ 3-4 independent meioses",Strength
408
+ SCN1A (HGNC:10585),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
409
+ SCN1A (HGNC:10585),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
410
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
411
+ SCN1A (HGNC:10585),PP3,Moderate,"Follow ClinGen’s recommendations (
412
+ PMID: 36413997
413
+ ) using REVEL as the computational tool, with the following stipulations:
414
+
415
+
416
+
417
+
418
+ Strength should be capped at Moderate, and 
419
+
420
+
421
+ limit the combination of PP3 and PM1 to reach no higher than strong","General recommendation,Strength"
422
+ SCN1A (HGNC:10585),PP3,Supporting,"Follow ClinGen’s recommendations (
423
+ PMID: 36413997
424
+ ) using REVEL as the computational tool, with the following stipulations: with the following stipulations:
425
+
426
+
427
+
428
+
429
+ Strength should be capped at Moderate, and 
430
+
431
+
432
+ limit the combination of PP3 and PM1 to reach no higher than strong","General recommendation,Strength"
433
+ SCN1A (HGNC:10585),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
434
+ SCN1A (HGNC:10585),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
435
+ SCN1A (HGNC:10585),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
436
+ SCN1A (HGNC:10585),BA1,Stand Alone,"Allele frequency is above 0.02% in GnomAD or other large population database, must be greater than or equal to 5 alleles if a minimum of 10,000 alleles was assessed.",Disease-specific
437
+ SCN1A (HGNC:10585),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
438
+ SCN1A (HGNC:10585),BS1,Strong,"Allele frequency is above 0.0004% in GnomAD or other large population database, must be greater than or equal to 5 alleles if a minimum of 10,000 alleles was assessed.",Disease-specific
439
+ SCN1A (HGNC:10585),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
440
+ SCN1A (HGNC:10585),BS2,Strong,Observed in a healthy adult individual.,No change
441
+ SCN1A (HGNC:10585),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,NA
442
+ SCN1A (HGNC:10585),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
443
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
444
+ SCN1A (HGNC:10585),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
445
+ SCN1A (HGNC:10585),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
446
+ SCN1A (HGNC:10585),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,No change
447
+ SCN1A (HGNC:10585),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,
448
+ SCN1A (HGNC:10585),BP3,Supporting,In frame-deletions/insertions in a repetitive region without a known function.,No change
449
+ SCN1A (HGNC:10585),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
450
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
451
+ SCN1A (HGNC:10585),BP4,Moderate,"Follow ClinGen’s recommendations (
452
+ PMID: 36413997
453
+ ) using REVEL as the computational tool.",General recommendation
454
+ SCN1A (HGNC:10585),BP4,Supporting,"Follow ClinGen’s recommendations (
455
+ PMID: 36413997
456
+ ) using REVEL as the computational tool.",General recommendation
457
+ SCN1A (HGNC:10585),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
458
+ SCN1A (HGNC:10585),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease.,No change
459
+ SCN1A (HGNC:10585),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
460
+ SCN1A (HGNC:10585),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
461
+ SCN1A (HGNC:10585),BP7,Supporting,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,No change
VCI/parsing_csr_criteria/version_csv_individual/ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1BVersion1.0.0_version=1.0.0.csv ADDED
@@ -0,0 +1,506 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ SCN1B (HGNC:10586),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ SCN1B (HGNC:10586),PVS1,Very Strong,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”, using SCN1B-specific information.
9
+
10
+
11
+ Most terminal codon predicted to undergo nonsense-mediated decay (NMD) p.Thr204. 
12
+
13
+
14
+ In-frame exons: 3-5
15
+
16
+
17
+ Out-of-frame exons: 1-2
18
+
19
+
20
+ For splice region variants, this criterion should not be applied with PP3.",General recommendation
21
+ SCN1B (HGNC:10586),PVS1,Strong,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”, using SCN1B-specific information.
22
+
23
+
24
+ Most terminal codon predicted to undergo nonsense-mediated decay (NMD) p.Thr204. 
25
+
26
+
27
+ In-frame exons: 3-5
28
+
29
+
30
+ Out-of-frame exons: 1-2
31
+
32
+
33
+ For splice region variants, this criterion should not be applied with PP3.",General recommendation
34
+ SCN1B (HGNC:10586),PVS1,Moderate,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”, using SCN1B-specific information.
35
+
36
+
37
+ Most terminal codon predicted to undergo nonsense-mediated decay (NMD) p.Thr204. 
38
+
39
+
40
+ In-frame exons: 3-5
41
+
42
+
43
+ Out-of-frame exons: 1-2
44
+
45
+
46
+ For splice region variants, this criterion should not be applied with PP3.",General recommendation
47
+ SCN1B (HGNC:10586),PVS1,Supporting,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”, using SCN1B-specific information.
48
+
49
+
50
+ Most terminal codon predicted to undergo nonsense-mediated decay (NMD) p.Thr204. 
51
+
52
+
53
+ In-frame exons: 3-5
54
+
55
+
56
+ Out-of-frame exons: 1-2
57
+
58
+
59
+ For splice region variants, this criterion should not be applied with PP3.",General recommendation
60
+ SCN1B (HGNC:10586),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
61
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
62
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
63
+ SCN1B (HGNC:10586),PS1,Strong,"Same amino acid change as a previously established
64
+ Pathogenic
65
+ variant regardless of nucleotide change. Example: Val->Leu caused by either G>C or G>T in the same codon. Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.
66
+
67
+
68
+ Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023).
69
+
70
+
71
+
72
+
73
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").",Gene-specific
74
+ SCN1B (HGNC:10586),PS1,Moderate,"Same amino acid change as a previously established
75
+ Likely Pathogenic
76
+ variant regardless of nucleotide change. Example: Val->Leu caused by either G>C or G>T in the same codon. Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.
77
+
78
+
79
+ Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023).
80
+
81
+
82
+
83
+
84
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").","Gene-specific,Strength"
85
+ SCN1B (HGNC:10586),PS1,Supporting,"Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023).
86
+
87
+
88
+
89
+
90
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").","Gene-specific,Strength"
91
+ SCN1B (HGNC:10586),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
92
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
93
+ SCN1B (HGNC:10586),PS2,Very Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
94
+ 4 points
95
+ will arrive at
96
+ Very Strong
97
+
98
+
99
+
100
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
101
+
102
+
103
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
104
+
105
+
106
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
107
+ SCN1B (HGNC:10586),PS2,Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
108
+ 2 points
109
+ will arrive at
110
+ Strong
111
+
112
+
113
+
114
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
115
+
116
+
117
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
118
+
119
+
120
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
121
+ SCN1B (HGNC:10586),PS2,Moderate,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
122
+ 1 point
123
+ will arrive at
124
+ Moderate
125
+
126
+
127
+
128
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
129
+
130
+
131
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
132
+
133
+
134
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
135
+ SCN1B (HGNC:10586),PS2,Supporting,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
136
+ 0.5 points
137
+ will arrive at
138
+ Supporting
139
+
140
+
141
+
142
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
143
+
144
+
145
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
146
+
147
+
148
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
149
+ SCN1B (HGNC:10586),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
150
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
151
+ SCN1B (HGNC:10586),PS3,Strong,Mouse knock-in model displays spontaneous seizures.,"Disease-specific,Gene-specific"
152
+ SCN1B (HGNC:10586),PS3,Moderate,"Heterologous expression with voltage clamping shows statistically significant difference over wildtype in at least one parameter (
153
+ https://bioportal.bioontology.org/ontologies/FENICS
154
+ )
155
+
156
+
157
+ Mouse knock-in model displays induced seizures
158
+
159
+
160
+ Zebrafish knock-in model displays spontaneous seizures, evidenced by hyperexcitability through electrophysiology or calcium imaging-based studies","Gene-specific,Strength"
161
+ SCN1B (HGNC:10586),PS3,Supporting,"Zebrafish knock-in model displays induced seizures, evidenced by hyperexcitability through electrophysiology or calcium imaging-based studies","Gene-specific,Strength"
162
+ SCN1B (HGNC:10586),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
163
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
164
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
165
+ SCN1B (HGNC:10586),PS4,Very Strong,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
166
+ 16+ points
167
+ will arrive at
168
+ Very Strong
169
+
170
+
171
+
172
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
173
+
174
+
175
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
176
+ SCN1B (HGNC:10586),PS4,Strong,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
177
+ 4-15.5 points
178
+ will arrive at
179
+ Strong
180
+
181
+
182
+
183
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
184
+
185
+
186
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific"
187
+ SCN1B (HGNC:10586),PS4,Moderate,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
188
+ 2-3.5 points
189
+ will arrive at
190
+ Moderate
191
+
192
+
193
+
194
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
195
+
196
+
197
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
198
+ SCN1B (HGNC:10586),PS4,Supporting,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
199
+ 1-1.5 points
200
+ will arrive at
201
+ Supporting
202
+
203
+
204
+
205
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
206
+
207
+
208
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
209
+ SCN1B (HGNC:10586),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
210
+ SCN1B (HGNC:10586),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
211
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
212
+ SCN1B (HGNC:10586),PM2,Supporting,"One or fewer alleles, if a minimum of 10,000 alleles assessed in population databases, such as the Genome Aggregation Database (gnomAD). Caveat: Population data for indels may be poorly called by next generation sequencing.","General recommendation,Other"
213
+ SCN1B (HGNC:10586),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
214
+ Note: This requires testing of parents (or offspring) to determine phase.",
215
+ SCN1B (HGNC:10586),PM3,Very Strong,"For recessive disorders, points-based system based on confirmation of phase and classification of other variant. Total of
216
+ 4.0 points
217
+ will arrive at
218
+ Very Strong
219
+
220
+
221
+
222
+
223
+
224
+ Classification of other variant is Pathogenic/Likely Pathogenic
225
+
226
+
227
+ Confirmed in trans: 1 point
228
+
229
+
230
+ Phase unknown: 0.5 (P) or 0.25 (LP) points
231
+
232
+
233
+
234
+
235
+
236
+
237
+ Homozygous occurrence (max point 1.0)
238
+
239
+
240
+ Confirmed in trans: 0.5 points
241
+
242
+
243
+
244
+
245
+
246
+
247
+ Classification of other variant is Uncertain Significance 
248
+
249
+
250
+ Confirmed in trans: 0.25 points
251
+
252
+
253
+ Phase unknown: 0 points",General recommendation
254
+ SCN1B (HGNC:10586),PM3,Strong,"For recessive disorders, points-based system based on confirmation of phase and classification of other variant. Total of
255
+ 2.0 points
256
+ will arrive at
257
+ Strong
258
+
259
+
260
+
261
+
262
+
263
+ Classification of other variant is Pathogenic/Likely Pathogenic
264
+
265
+
266
+ Confirmed in trans: 1 point
267
+
268
+
269
+ Phase unknown: 0.5 (P) or 0.25 (LP) points
270
+
271
+
272
+
273
+
274
+
275
+
276
+ Homozygous occurrence (max point 1.0)
277
+
278
+
279
+ Confirmed in trans: 0.5 points
280
+
281
+
282
+
283
+
284
+
285
+
286
+ Classification of other variant is Uncertain Significance 
287
+
288
+
289
+ Confirmed in trans: 0.25 points
290
+
291
+
292
+ Phase unknown: 0 points",General recommendation
293
+ SCN1B (HGNC:10586),PM3,Moderate,"For recessive disorders, points-based system based on confirmation of phase and classification of other variant. Total of
294
+ 1.0 point
295
+ will arrive at
296
+ Moderate
297
+
298
+
299
+
300
+
301
+
302
+ Classification of other variant is Pathogenic/Likely Pathogenic
303
+
304
+
305
+ Confirmed in trans: 1 point
306
+
307
+
308
+ Phase unknown: 0.5 (P) or 0.25 (LP) points
309
+
310
+
311
+
312
+
313
+
314
+
315
+ Homozygous occurrence (max point 1.0)
316
+
317
+
318
+ Confirmed in trans: 0.5 points
319
+
320
+
321
+
322
+
323
+
324
+
325
+ Classification of other variant is Uncertain Significance 
326
+
327
+
328
+ Confirmed in trans: 0.25 points
329
+
330
+
331
+ Phase unknown: 0 points",General recommendation
332
+ SCN1B (HGNC:10586),PM3,Supporting,"For recessive disorders, points-based system based on confirmation of phase and classification of other variant. Total of
333
+ 0.5 points
334
+ will arrive at
335
+ Supporting
336
+
337
+
338
+
339
+
340
+
341
+ Classification of other variant is Pathogenic/Likely Pathogenic
342
+
343
+
344
+ Confirmed in trans: 1 point
345
+
346
+
347
+ Phase unknown: 0.5 (P) or 0.25 (LP) points
348
+
349
+
350
+
351
+
352
+
353
+
354
+ Homozygous occurrence (max point 1.0)
355
+
356
+
357
+ Confirmed in trans: 0.5 points
358
+
359
+
360
+
361
+
362
+
363
+
364
+ Classification of other variant is Uncertain Significance 
365
+
366
+
367
+ Confirmed in trans: 0.25 points
368
+
369
+
370
+ Phase unknown: 0 points",General recommendation
371
+ SCN1B (HGNC:10586),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
372
+ SCN1B (HGNC:10586),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
373
+ SCN1B (HGNC:10586),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
374
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
375
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
376
+ SCN1B (HGNC:10586),PM5,Strong,This should say greater than or equal to 2 known pathogenic variants at same site as novel change.,"Disease-specific,Gene-specific,Strength"
377
+ SCN1B (HGNC:10586),PM5,Moderate,"Novel missense change at an amino acid residue where a different missense change determined to be
378
+ Pathogenic
379
+ has been seen before. Example: Arg156His is pathogenic; now you observe Arg156Cys.
380
+
381
+
382
+
383
+
384
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.","Disease-specific,Gene-specific"
385
+ SCN1B (HGNC:10586),PM5,Supporting,"Novel missense change at an amino acid residue where a different missense change determined to be
386
+ Likely Pathogenic
387
+ has been seen before. Example: Arg156His is pathogenic; now you observe Arg156Cys.","Disease-specific,Gene-specific,Strength"
388
+ SCN1B (HGNC:10586),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
389
+ SCN1B (HGNC:10586),PM6,Strong,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
390
+ 2 points
391
+ will arrive at
392
+ Strong
393
+ . Total of
394
+ 4 points
395
+ will arrive at
396
+ Very Strong
397
+
398
+
399
+
400
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 0.5 points
401
+
402
+
403
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
404
+ SCN1B (HGNC:10586),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
405
+ 1 point
406
+ will arrive at
407
+ Moderate
408
+
409
+
410
+
411
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 0.5 points
412
+
413
+
414
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
415
+ SCN1B (HGNC:10586),PM6,Supporting,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
416
+ 0.5 points
417
+ will arrive at
418
+ Supporting
419
+
420
+
421
+
422
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 0.5 points
423
+
424
+
425
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
426
+ SCN1B (HGNC:10586),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
427
+ Note: May be used as stronger evidence with increasing segregation data.",
428
+ SCN1B (HGNC:10586),PP1,Strong,"Co-segregation with disease in multiple affected family members
429
+
430
+
431
+ AD: ≥7 independent meioses
432
+
433
+
434
+ AR: ≥3 affected segregations",Strength
435
+ SCN1B (HGNC:10586),PP1,Moderate,"Co-segregation with disease in multiple affected family members
436
+
437
+
438
+ AD: 5-6 independent meioses
439
+
440
+
441
+ AR: 2 affected segregations",Strength
442
+ SCN1B (HGNC:10586),PP1,Supporting,"Co-segregation with disease in multiple affected family members
443
+
444
+
445
+ AD: 3-4 independent meioses
446
+
447
+
448
+ AR: 1 affected segregation",Strength
449
+ SCN1B (HGNC:10586),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
450
+ SCN1B (HGNC:10586),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
451
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
452
+ SCN1B (HGNC:10586),PP3,Moderate,"Follow ClinGen’s recommendations (
453
+ PMID: 36413997
454
+ ), using REVEL as the computational tool, with the following stipulations:
455
+
456
+
457
+
458
+
459
+ Strength should be capped at Moderate, and 
460
+
461
+
462
+ limit the combination of PP3 and PM1 to reach no higher than strong","General recommendation,Strength"
463
+ SCN1B (HGNC:10586),PP3,Supporting,"Follow ClinGen’s recommendations (
464
+ PMID: 36413997
465
+ ), using REVEL as the computational tool, with the following stipulations:
466
+
467
+
468
+
469
+
470
+ Strength should be capped at Moderate, and 
471
+
472
+
473
+ limit the combination of PP3 and PM1 to reach no higher than strong","General recommendation,Strength"
474
+ SCN1B (HGNC:10586),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
475
+ SCN1B (HGNC:10586),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
476
+ SCN1B (HGNC:10586),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
477
+ SCN1B (HGNC:10586),BA1,Stand Alone,"Allele frequency is above
478
+ 0.3%
479
+ in GnomAD or other large population database, must be greater than or equal to 5 alleles if a minimum of 10,000 alleles was assessed.",Disease-specific
480
+ SCN1B (HGNC:10586),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
481
+ SCN1B (HGNC:10586),BS1,Strong,"Allele frequency is above
482
+ 0.01%
483
+ in GnomAD or other large population database, must be greater than or equal to 5 alleles if a minimum of 10,000 alleles was assessed.",Disease-specific
484
+ SCN1B (HGNC:10586),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
485
+ SCN1B (HGNC:10586),BS2,Strong,Observed in a healthy adult individual.,No change
486
+ SCN1B (HGNC:10586),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,NA
487
+ SCN1B (HGNC:10586),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
488
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
489
+ SCN1B (HGNC:10586),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
490
+ SCN1B (HGNC:10586),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
491
+ SCN1B (HGNC:10586),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,No change
492
+ SCN1B (HGNC:10586),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,
493
+ SCN1B (HGNC:10586),BP3,Supporting,In frame-deletions/insertions in a repetitive region without a known function.,No change
494
+ SCN1B (HGNC:10586),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
495
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
496
+ SCN1B (HGNC:10586),BP4,Moderate,"Follow ClinGen’s recommendations (
497
+ PMID: 36413997
498
+ ), using REVEL as the computational tool.",General recommendation
499
+ SCN1B (HGNC:10586),BP4,Supporting,"Follow ClinGen’s recommendations (
500
+ PMID: 36413997
501
+ ), using REVEL as the computational tool.",General recommendation
502
+ SCN1B (HGNC:10586),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
503
+ SCN1B (HGNC:10586),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease.,No change
504
+ SCN1B (HGNC:10586),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
505
+ SCN1B (HGNC:10586),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
506
+ SCN1B (HGNC:10586),BP7,Supporting,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,No change
VCI/parsing_csr_criteria/version_csv_individual/ClinGenEpilepsySodiumChannelExpertPanelSpecificationstotheACMGAMPVariantInterpretationGuidelinesforSCN1BVersion2.0.0_version=2.0.0.csv ADDED
@@ -0,0 +1,458 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ Gene,Code,Strength,Description,Modification Type
2
+ SCN1B (HGNC:10586),PVS1,Original ACMG Summary,"Null variant (nonsense, frameshift, canonical +/−1 or 2 splice sites, initiation codon, single or multi-exon deletion) in a gene where loss of function (LOF) is a known mechanism of disease.
3
+ Caveats:
4
+  • Beware of genes where LOF is not a known disease mechanism (e.g. GFAP, MYH7).
5
+  • Use caution interpreting LOF variants at the extreme 3’ end of a gene.
6
+  • Use caution with splice variants that are predicted to lead to exon skipping but leave the remainder of the protein intact.
7
+  • Use caution in the presence of multiple transcripts.",
8
+ SCN1B (HGNC:10586),PVS1,Very Strong,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”.",General recommendation
9
+ SCN1B (HGNC:10586),PVS1,Strong,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”.",General recommendation
10
+ SCN1B (HGNC:10586),PVS1,Moderate,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”.",General recommendation
11
+ SCN1B (HGNC:10586),PVS1,Supporting,"Follow SVI guidance per workflow in Tayoun et al (2018), included as “PVS1 Decision Tree”.",General recommendation
12
+ SCN1B (HGNC:10586),PS1,Original ACMG Summary,"Same amino acid change as a previously established pathogenic variant regardless of nucleotide change.
13
+ Example: Val->Leu caused by either G>C or G>T in the same codon.
14
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
15
+ SCN1B (HGNC:10586),PS1,Strong,"Same amino acid change as a previously established
16
+ Pathogenic
17
+ variant regardless of nucleotide change. Example: Val->Leu caused by either G>C or G>T in the same codon. Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.
18
+
19
+
20
+ Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023).
21
+
22
+
23
+
24
+
25
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").",Gene-specific
26
+ SCN1B (HGNC:10586),PS1,Moderate,"Same amino acid change as a previously established
27
+ Likely Pathogenic
28
+ variant regardless of nucleotide change. Example: Val->Leu caused by either G>C or G>T in the same codon. Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.
29
+
30
+
31
+ Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023).
32
+
33
+
34
+
35
+
36
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").","Gene-specific,Strength"
37
+ SCN1B (HGNC:10586),PS1,Supporting,"Same predicted impact on splicing as previously classified variant (Refer to Table 2 in Walker et al, 2023).
38
+
39
+
40
+
41
+
42
+ PS1 can be applied at varying strengths for splice variants, in conjunction with either PP3 or PVS1. PS1 strength depends on location of the variant under assessment (within or outside the +/- 1,2 dinucleotide positions) and the location of the previously classified variant (within or outside the +/- 1,2 dinucleotide position). Specific combinations are outlined in Table 2 in Walker, et al (2023) PMID: 37352859, also provided as a supplement (""PS1_Variants impacting splicing"").","Gene-specific,Strength"
43
+ SCN1B (HGNC:10586),PS2,Original ACMG Summary,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history.
44
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.",
45
+ SCN1B (HGNC:10586),PS2,Very Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
46
+ 4 points
47
+ will arrive at
48
+ Very Strong
49
+
50
+
51
+
52
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
53
+
54
+
55
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
56
+
57
+
58
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
59
+ SCN1B (HGNC:10586),PS2,Strong,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
60
+ 2 points
61
+ will arrive at
62
+ Strong
63
+
64
+
65
+
66
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
67
+
68
+
69
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
70
+
71
+
72
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
73
+ SCN1B (HGNC:10586),PS2,Moderate,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
74
+ 1 point
75
+ will arrive at
76
+ Moderate
77
+
78
+
79
+
80
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
81
+
82
+
83
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
84
+
85
+
86
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
87
+ SCN1B (HGNC:10586),PS2,Supporting,"De novo (both maternity and paternity confirmed) in a patient with the disease and no family history. Points based system for each unrelated proband determined by phenotypic specificity. Total of
88
+ 0.5 points
89
+ will arrive at
90
+ Supporting
91
+
92
+
93
+
94
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
95
+
96
+
97
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points
98
+
99
+
100
+ Note: Confirmation of paternity only is insufficient. Egg donation, surrogate motherhood, errors in embryo transfer, etc. can contribute to non-maternity.","Disease-specific,Strength"
101
+ SCN1B (HGNC:10586),PS3,Original ACMG Summary,"Well-established in vitro or in vivo functional studies supportive of a damaging effect on the gene or gene product.
102
+ Note: Functional studies that have been validated and shown to be reproducible and robust in a clinical diagnostic laboratory setting are considered the most well-established.",
103
+ SCN1B (HGNC:10586),PS3,Strong,Mouse knock-in model displays spontaneous seizures.,"Disease-specific,Gene-specific"
104
+ SCN1B (HGNC:10586),PS3,Moderate,"Heterologous expression with voltage clamping shows statistically significant difference over wildtype in at least one parameter (
105
+ https://bioportal.bioontology.org/ontologies/FENICS
106
+ )
107
+
108
+
109
+ Mouse knock-in model displays induced seizures
110
+
111
+
112
+ Zebrafish knock-in model displays spontaneous seizures, evidenced by hyperexcitability through electrophysiology or calcium imaging-based studies","Gene-specific,Strength"
113
+ SCN1B (HGNC:10586),PS3,Supporting,"Zebrafish knock-in model displays induced seizures, evidenced by hyperexcitability through electrophysiology or calcium imaging-based studies","Gene-specific,Strength"
114
+ SCN1B (HGNC:10586),PS4,Original ACMG Summary,"The prevalence of the variant in affected individuals is significantly increased compared to the prevalence in controls.
115
+ Note 1: Relative risk (RR) or odds ratio (OR), as obtained from case-control studies, is >5.0 and the confidence interval around the estimate of RR or OR does not include 1.0. See manuscript for detailed guidance.
116
+ Note 2: In instances of very rare variants where case-control studies may not reach statistical significance, the prior observation of the variant in multiple unrelated patients with the same phenotype, and its absence in controls, may be used as moderate level of evidence.",
117
+ SCN1B (HGNC:10586),PS4,Very Strong,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
118
+ 16+ points
119
+ will arrive at
120
+ Very Strong
121
+
122
+
123
+
124
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
125
+
126
+
127
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
128
+ SCN1B (HGNC:10586),PS4,Strong,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
129
+ 4-15.5 points
130
+ will arrive at
131
+ Strong
132
+
133
+
134
+
135
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
136
+
137
+
138
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific"
139
+ SCN1B (HGNC:10586),PS4,Moderate,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
140
+ 2-3.5 points
141
+ will arrive at
142
+ Moderate
143
+
144
+
145
+
146
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
147
+
148
+
149
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
150
+ SCN1B (HGNC:10586),PS4,Supporting,"Present in in multiple unrelated patients with consistent phenotype. Points based system for each unrelated proband determined by phenotypic specificity. Total of
151
+ 1-1.5 points
152
+ will arrive at
153
+ Supporting
154
+
155
+
156
+
157
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 1 point
158
+
159
+
160
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.5 points","Disease-specific,Gene-specific,Strength"
161
+ SCN1B (HGNC:10586),PM1,Original ACMG Summary,Located in a mutational hot spot and/or critical and well-established functional domain (e.g. active site of an enzyme) without benign variation.,NA
162
+ SCN1B (HGNC:10586),PM2,Original ACMG Summary,"Absent from controls (or at extremely low frequency if recessive) in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.
163
+ Caveat: Population data for indels may be poorly called by next generation sequencing.",
164
+ SCN1B (HGNC:10586),PM2,Supporting,"One or fewer alleles, if a minimum of 10,000 alleles assessed in population databases, such as the Genome Aggregation Database (gnomAD). Caveat: Population data for indels may be poorly called by next generation sequencing.","General recommendation,Other"
165
+ SCN1B (HGNC:10586),PM3,Original ACMG Summary,"For recessive disorders, detected in trans with a pathogenic variant
166
+ Note: This requires testing of parents (or offspring) to determine phase.",
167
+ SCN1B (HGNC:10586),PM3,Very Strong,"For recessive disorders, points-based system based on confirmation of phase and classification of other variant. Total of
168
+ 4.0 points
169
+ will arrive at
170
+ Very Strong
171
+
172
+
173
+
174
+
175
+
176
+ Classification of other variant is Pathogenic/Likely Pathogenic
177
+
178
+
179
+ Confirmed in trans: 1 point
180
+
181
+
182
+ Phase unknown: 0.5 (P) or 0.25 (LP) points
183
+
184
+
185
+
186
+
187
+
188
+
189
+ Homozygous occurrence (max point 1.0)
190
+
191
+
192
+ Confirmed in trans: 0.5 points
193
+
194
+
195
+
196
+
197
+
198
+
199
+ Classification of other variant is Uncertain Significance 
200
+
201
+
202
+ Confirmed in trans: 0.25 points
203
+
204
+
205
+ Phase unknown: 0 points",General recommendation
206
+ SCN1B (HGNC:10586),PM3,Strong,"For recessive disorders, points-based system based on confirmation of phase and classification of other variant. Total of
207
+ 2.0 points
208
+ will arrive at
209
+ Strong
210
+
211
+
212
+
213
+
214
+
215
+ Classification of other variant is Pathogenic/Likely Pathogenic
216
+
217
+
218
+ Confirmed in trans: 1 point
219
+
220
+
221
+ Phase unknown: 0.5 (P) or 0.25 (LP) points
222
+
223
+
224
+
225
+
226
+
227
+
228
+ Homozygous occurrence (max point 1.0)
229
+
230
+
231
+ Confirmed in trans: 0.5 points
232
+
233
+
234
+
235
+
236
+
237
+
238
+ Classification of other variant is Uncertain Significance 
239
+
240
+
241
+ Confirmed in trans: 0.25 points
242
+
243
+
244
+ Phase unknown: 0 points",General recommendation
245
+ SCN1B (HGNC:10586),PM3,Moderate,"For recessive disorders, points-based system based on confirmation of phase and classification of other variant. Total of
246
+ 1.0 point
247
+ will arrive at
248
+ Moderate
249
+
250
+
251
+
252
+
253
+
254
+ Classification of other variant is Pathogenic/Likely Pathogenic
255
+
256
+
257
+ Confirmed in trans: 1 point
258
+
259
+
260
+ Phase unknown: 0.5 (P) or 0.25 (LP) points
261
+
262
+
263
+
264
+
265
+
266
+
267
+ Homozygous occurrence (max point 1.0)
268
+
269
+
270
+ Confirmed in trans: 0.5 points
271
+
272
+
273
+
274
+
275
+
276
+
277
+ Classification of other variant is Uncertain Significance 
278
+
279
+
280
+ Confirmed in trans: 0.25 points
281
+
282
+
283
+ Phase unknown: 0 points",General recommendation
284
+ SCN1B (HGNC:10586),PM3,Supporting,"For recessive disorders, points-based system based on confirmation of phase and classification of other variant. Total of
285
+ 0.5 points
286
+ will arrive at
287
+ Supporting
288
+
289
+
290
+
291
+
292
+
293
+ Classification of other variant is Pathogenic/Likely Pathogenic
294
+
295
+
296
+ Confirmed in trans: 1 point
297
+
298
+
299
+ Phase unknown: 0.5 (P) or 0.25 (LP) points
300
+
301
+
302
+
303
+
304
+
305
+
306
+ Homozygous occurrence (max point 1.0)
307
+
308
+
309
+ Confirmed in trans: 0.5 points
310
+
311
+
312
+
313
+
314
+
315
+
316
+ Classification of other variant is Uncertain Significance 
317
+
318
+
319
+ Confirmed in trans: 0.25 points
320
+
321
+
322
+ Phase unknown: 0 points",General recommendation
323
+ SCN1B (HGNC:10586),PM4,Original ACMG Summary,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,
324
+ SCN1B (HGNC:10586),PM4,Moderate,Protein length changes due to in-frame deletions/insertions in a non-repeat region or stop-loss variants.,No change
325
+ SCN1B (HGNC:10586),PM5,Original ACMG Summary,"Novel missense change at an amino acid residue where a different missense change determined to be pathogenic has been seen before.
326
+ Example: Arg156His is pathogenic; now you observe Arg156Cys.
327
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.",
328
+ SCN1B (HGNC:10586),PM5,Strong,This should say greater than or equal to 2 known pathogenic variants at same site as novel change.,"Disease-specific,Gene-specific,Strength"
329
+ SCN1B (HGNC:10586),PM5,Moderate,"Novel missense change at an amino acid residue where a different missense change determined to be
330
+ Pathogenic
331
+ has been seen before. Example: Arg156His is pathogenic; now you observe Arg156Cys.
332
+
333
+
334
+
335
+
336
+ Caveat: Beware of changes that impact splicing rather than at the amino acid/protein level.","Disease-specific,Gene-specific"
337
+ SCN1B (HGNC:10586),PM5,Supporting,"Novel missense change at an amino acid residue where a different missense change determined to be
338
+ Likely Pathogenic
339
+ has been seen before. Example: Arg156His is pathogenic; now you observe Arg156Cys.","Disease-specific,Gene-specific,Strength"
340
+ SCN1B (HGNC:10586),PM6,Original ACMG Summary,"Assumed de novo, but without confirmation of paternity and maternity.",
341
+ SCN1B (HGNC:10586),PM6,Strong,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
342
+ 2 points
343
+ will arrive at
344
+ Strong
345
+ . Total of
346
+ 4 points
347
+ will arrive at
348
+ Very Strong
349
+
350
+
351
+
352
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 0.5 points
353
+
354
+
355
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
356
+ SCN1B (HGNC:10586),PM6,Moderate,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
357
+ 1 point
358
+ will arrive at
359
+ Moderate
360
+
361
+
362
+
363
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 0.5 points
364
+
365
+
366
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
367
+ SCN1B (HGNC:10586),PM6,Supporting,"Assumed de novo, but without confirmation of paternity and maternity. Points based system for each unrelated proband determined by phenotypic specificity. Total of
368
+ 0.5 points
369
+ will arrive at
370
+ Supporting
371
+
372
+
373
+
374
+ Genetic Epilepsy with Febrile Seizures Plus (GEFS+): 0.5 points
375
+
376
+
377
+ Other epilepsy types or syndromes not included above, with or without associated neurodevelopmental features: 0.25 points","Disease-specific,Strength"
378
+ SCN1B (HGNC:10586),PP1,Original ACMG Summary,"Co-segregation with disease in multiple affected family members in a gene definitively known to cause the disease.
379
+ Note: May be used as stronger evidence with increasing segregation data.",
380
+ SCN1B (HGNC:10586),PP1,Strong,"Co-segregation with disease in multiple affected family members
381
+
382
+
383
+ AD: ≥7 independent meioses
384
+
385
+
386
+ AR: ≥3 affected segregations",Strength
387
+ SCN1B (HGNC:10586),PP1,Moderate,"Co-segregation with disease in multiple affected family members
388
+
389
+
390
+ AD: 5-6 independent meioses
391
+
392
+
393
+ AR: 2 affected segregations",Strength
394
+ SCN1B (HGNC:10586),PP1,Supporting,"Co-segregation with disease in multiple affected family members
395
+
396
+
397
+ AD: 3-4 independent meioses
398
+
399
+
400
+ AR: 1 affected segregation",Strength
401
+ SCN1B (HGNC:10586),PP2,Original ACMG Summary,Missense variant in a gene that has a low rate of benign missense variation and where missense variants are a common mechanism of disease.,NA
402
+ SCN1B (HGNC:10586),PP3,Original ACMG Summary,"Multiple lines of computational evidence support a deleterious effect on the gene or gene product (conservation, evolutionary, splicing impact, etc.).
403
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm should not be counted as an independent criterion. PP3 can be used only once in any evaluation of a variant.",
404
+ SCN1B (HGNC:10586),PP3,Moderate,"Follow ClinGen’s recommendations (
405
+ PMID: 36413997
406
+ ), using REVEL as the computational tool, with the following stipulations:
407
+
408
+
409
+
410
+
411
+ Strength should be capped at Moderate, and 
412
+
413
+
414
+ limit the combination of PP3 and PM1 to reach no higher than strong","General recommendation,Strength"
415
+ SCN1B (HGNC:10586),PP3,Supporting,"Follow ClinGen’s recommendations (
416
+ PMID: 36413997
417
+ ), using REVEL as the computational tool, with the following stipulations:
418
+
419
+
420
+
421
+
422
+ Strength should be capped at Moderate, and 
423
+
424
+
425
+ limit the combination of PP3 and PM1 to reach no higher than strong","General recommendation,Strength"
426
+ SCN1B (HGNC:10586),PP4,Original ACMG Summary,Patient’s phenotype or family history is highly specific for a disease with a single genetic etiology.,NA
427
+ SCN1B (HGNC:10586),PP5,Original ACMG Summary,"Reputable source recently reports variant as pathogenic, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
428
+ SCN1B (HGNC:10586),BA1,Original ACMG Summary,"Allele frequency is above 5% in Exome Sequencing Project, 1000 Genomes or Exome Aggregation Consortium.",
429
+ SCN1B (HGNC:10586),BA1,Stand Alone,"Allele frequency is above
430
+ 0.3%
431
+ in GnomAD or other large population database, must be greater than or equal to 5 alleles if a minimum of 10,000 alleles was assessed.",Disease-specific
432
+ SCN1B (HGNC:10586),BS1,Original ACMG Summary,Allele frequency is greater than expected for disorder.,
433
+ SCN1B (HGNC:10586),BS1,Strong,"Allele frequency is above
434
+ 0.01%
435
+ in GnomAD or other large population database, must be greater than or equal to 5 alleles if a minimum of 10,000 alleles was assessed.",Disease-specific
436
+ SCN1B (HGNC:10586),BS2,Original ACMG Summary,"Observed in a healthy adult individual for a recessive (homozygous), dominant (heterozygous), or X-linked (hemizygous) disorder, with full penetrance expected at an early age.",
437
+ SCN1B (HGNC:10586),BS2,Strong,Observed in a healthy adult individual.,No change
438
+ SCN1B (HGNC:10586),BS3,Original ACMG Summary,Well-established in vitro or in vivo functional studies show no damaging effect on protein function or splicing.,NA
439
+ SCN1B (HGNC:10586),BS4,Original ACMG Summary,"Lack of segregation in affected members of a family.
440
+ Caveat: The presence of phenocopies for common phenotypes (i.e. cancer, epilepsy) can mimic lack of segregation among affected individuals. Also, families may have more than one pathogenic variant contributing to an autosomal dominant disorder, further confounding an apparent lack of segregation.",NA
441
+ SCN1B (HGNC:10586),BP1,Original ACMG Summary,Missense variant in a gene for which primarily truncating variants are known to cause disease.,NA
442
+ SCN1B (HGNC:10586),BP2,Original ACMG Summary,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,
443
+ SCN1B (HGNC:10586),BP2,Supporting,Observed in trans with a pathogenic variant for a fully penetrant dominant gene/disorder or observed in cis with a pathogenic variant in any inheritance pattern.,No change
444
+ SCN1B (HGNC:10586),BP3,Original ACMG Summary,In frame-deletions/insertions in a repetitive region without a known function.,
445
+ SCN1B (HGNC:10586),BP3,Supporting,In frame-deletions/insertions in a repetitive region without a known function.,No change
446
+ SCN1B (HGNC:10586),BP4,Original ACMG Summary,"Multiple lines of computational evidence suggest no impact on gene or gene product (conservation, evolutionary, splicing impact, etc)
447
+ Caveat: As many in silico algorithms use the same or very similar input for their predictions, each algorithm cannot be counted as an independent criterion. BP4 can be used only once in any evaluation of a variant.",
448
+ SCN1B (HGNC:10586),BP4,Moderate,"Follow ClinGen’s recommendations (
449
+ PMID: 36413997
450
+ ), using REVEL as the computational tool.",General recommendation
451
+ SCN1B (HGNC:10586),BP4,Supporting,"Follow ClinGen’s recommendations (
452
+ PMID: 36413997
453
+ ), using REVEL as the computational tool.",General recommendation
454
+ SCN1B (HGNC:10586),BP5,Original ACMG Summary,Variant found in a case with an alternate molecular basis for disease.,
455
+ SCN1B (HGNC:10586),BP5,Supporting,Variant found in a case with an alternate molecular basis for disease.,No change
456
+ SCN1B (HGNC:10586),BP6,Original ACMG Summary,"Reputable source recently reports variant as benign, but the evidence is not available to the laboratory to perform an independent evaluation.",NA
457
+ SCN1B (HGNC:10586),BP7,Original ACMG Summary,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,
458
+ SCN1B (HGNC:10586),BP7,Supporting,A synonymous variant for which splicing prediction algorithms predict no impact to the splice consensus sequence nor the creation of a new splice site AND the nucleotide is not highly conserved.,No change