neoantigens: the Sahin TNBC vaccine cohort (Nature 2025, s41586-025-10004-2)

251 vaccinated targets over 14 patients, from supplementary sheet "n=251 targets". Ex-vivo
ELISpot: 43 positive (41 de novo + 2 amplified) against 208 no-response, 17.1% prevalence.
Post-IVS best call on the 85 targets with data: 44 positive, 41 negative. mean_spot_count is
the measured continuous magnitude, kept because a binary call discards it.

TARGET-LEVEL, not peptide-level, and that is the caveat to read before joining it to anything:
only 8 of 251 rows carry a minimal epitope. The remaining 243 are a gene plus a substitution,
and expanding them to (peptide, allele) needs the patients' HLA typing -- which is NOT
published in either the paper or its supplements. Only 7 distinct alleles appear anywhere in
them, and those are the CD8 restrictions of the identified TCRs in P1/P12/P13, not a typing
table. Extended Data Fig. 4 is the Jurkat-NFAT TCR validation figure.

peptide_level flags which rows are directly usable. A response call that cannot be read is
recorded as missing, never as a negative.

neoantigens/sahin_tnbc_2025.tsv.gz

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:bce12324657f0a326bd8189a10194972b83f3bf9e995744c05f10c47ad1c942d
+size 4697