First-time data synchronization of GitHub directory and data

CHANGELOG.md

@@ -0,0 +1,22 @@
+# Changelog
+
+All notable changes to this project will be documented in this file.  
+This project follows semantic versioning where possible.
+
+---
+
+## [v0.1] - 2025-09-24
+### Added
+- Initial public release of the **Neoantigen Discovery Dataset (NDD)**.
+- Main dataset file `ndd_v0.1.tsv` containing curated **basic fields** and **predicted features**.
+- Patient-level metadata file `patient_metadata.tsv` with limited cohort information.
+- Documentation under `docs/`:
+  - `field-schema.md` — definitions of basic fields
+  - `predicted-features.md` — descriptions of predicted features and tools
+  - `patient-metadata.md` — patient-level metadata definitions
+- Dual license:
+  - **MIT** for code/scripts
+  - **CC-BY 4.0** for dataset files
+
+---
+

DATA_LICENSE

@@ -0,0 +1,13 @@
+Creative Commons Attribution 4.0 International (CC BY 4.0)
+
+This dataset is licensed under the Creative Commons Attribution 4.0 International License.
+
+You are free to:
+- Share — copy and redistribute the material in any medium or format
+- Adapt — remix, transform, and build upon the material for any purpose, even commercially.
+
+Under the following terms:
+- Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. 
+  You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.
+
+Full license text: https://creativecommons.org/licenses/by/4.0/

LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2025 Neoantigen Discovery Dataset (NDD) Contributors
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

README.md

@@ -1,3 +1,171 @@
----
-license: cc-by-4.0
----
+# Neoantigen Discovery Dataset (NDD) - v0.1
+
+---
+
+## Overview
+
+The Neoantigen Discovery Dataset (NDD) was developed with a clear focus on machine learning applications in neoantigen prediction. From the outset, NDD has been designed to support model development by systematically collecting a broad range of dimensions — including basic information, experimental labels, and ML-ready predictive features that can be directly used for training and validation.
+
+In neoantigen research, different strategies serve complementary roles: threshold-based approaches are effective for filtering candidates using known rules, while deep learning methods require large-scale, standardized datasets to uncover complex patterns and improve generalization.
+
+NDD addresses this need by providing a curated, standardized, and feature-rich dataset that enables the development, training, and benchmarking of new algorithms, while also facilitating integration with experimental evidence.
+
+### Key highlights include
+
+- Integration of data from literature reports and public resources such as IEDB, TCGA, and CEDAR.  
+- Coverage of core experimental evidence alongside predicted features (binding affinity, stability, anchor mutations, driver status, etc.).  
+- A structured design that supports both peptide-level predictions and patient-level cohort analyses.  
+- Preservation of assay details to ensure reliable immunogenicity labels and reproducibility.  
+
+NDD is not just a file collection, but a community resource for advancing model development, evaluation, and integration with experimental data. The current release, NDD v0.1, marks the starting point of an ongoing effort to expand data coverage, enrich feature annotations, and promote collaboration across the research community.
+
+---
+
+## Data Sources & Curation
+
+NDD entries are primarily curated from original literature reports, manually reviewed and standardized for consistency. To ensure transparency and traceability, each record includes PMID/DOI identifiers, allowing users to directly access the corresponding publications.
+
+In addition, we drew inspiration from previously published neoantigen datasets summarized by other research groups, building on their foundational efforts while extending the scope through broader standardization and feature integration.
+
+The curation process draws on the frameworks established by public databases such as IEDB and CEDAR, while also integrating cross-references to resources like TCGA. This approach ensures that NDD combines the breadth of literature-derived evidence with the rigor of structured annotation.
+
+### Standardization Principles
+
+- Mutation nomenclature: Reported using HGVS format (e.g., KRAS p.G12D). Mutation descriptions are preserved as given in the original publications, which may reference different genome builds (commonly hg19 or hg38). In v0.1, all entries retain their original reporting style; future versions will unify genome references to improve comparability.  
+- Cancer type classification: Manually normalized and mapped to widely used taxonomies such as IntOGen and TCGA, ensuring consistent categorization across diverse studies.  
+- HLA typing: Harmonized to four-digit resolution (e.g., HLA-A02:01), with aliases and shorthand notations consolidated.  
+
+### Quality Control
+
+- Missing values are explicitly marked as NA.  
+- Consistency checks combine manual review with automated scripts to validate key fields.  
+- Future improvements will include automated QC pipelines to further minimize potential errors.  
+- Source traceability is preserved for every entry, ensuring reproducibility and enabling direct comparison with original publications.  
+
+Detailed documentation of the tools, versions, and parameters used for feature annotation is provided in `docs/predicted-features.md`.
+
+---
+
+## Fields & Features
+
+NDD provides information at three levels: basic fields, predicted features, and patient-level metadata. Together, these dimensions make the dataset suitable for modeling and exploratory research.
+
+- **Basic fields** capture essential information from original reports, including:  
+  patient ID, tumor type, gene/mutation (HGVS), mutant and wild-type peptides, HLA allele, experimental method.  
+  These fields form the foundation of each entry.  
+
+- **Predicted features** extend each record with annotations widely used in neoantigen research, such as:  
+  binding affinity and rank (NetMHCpan), stability (NetMHCstabpan), recognition likelihood (PRIME, BigMHC_IM), cleavage and transport scores (NetChop, NetCTLpan), differential agretopicity index (DAI), anchor mutation position, TCGA cancer expression (TPM median), and driver gene status (IntOGen).  
+  These features make the dataset directly applicable to machine learning tasks.  
+
+- **Patient-level metadata** (when available) include additional information such as extended HLA typing, clinical outcomes, or treatment details. In the current release these data are limited and not yet complete, but we recognize their importance and will continue to expand and refine this part of the dataset in future versions.  
+
+**Field definitions and detailed documentation are provided in the `docs/` directory, including:**  
+- `field-schema.md` — definitions of basic fields  
+- `predicted-features.md` — list and description of predicted features  
+- `patient-metadata.md` — definitions of patient-level metadata (when available)  
+
+---
+
+## Usage & Format
+
+The dataset is released in tab-separated values (TSV, UTF-8 encoded) format.
+
+- **Main file:** `data/ndd_v0.1.tsv`  
+  Each row represents a mutation–HLA–peptide entry, combining both basic fields and predicted features.  
+
+- **Patient-level metadata:** `data/patient_metadata.tsv`  
+  Contains information at the patient dimension, such as extended HLA typing, clinical outcomes, and treatment details when available.  
+
+*Note:* Patient IDs are anonymized to ensure privacy but remain consistent across files, enabling cohort-level analysis.  
+
+### Limitations
+
+- Patient-level metadata are currently limited in number and completeness, reflecting what is available in published sources.  
+- Certain molecular features (e.g., RNA-seq coverage, expression levels, clonality, cancer cell fraction) are not yet integrated in this release. Only TCGA cancer expression (TPM median) is included.  
+- Cancer type mapping to TCGA or IntOGen was performed internally during curation but is not included in the public dataset. Researchers may apply their own mapping according to their needs.  
+
+These reflect the present scope of published evidence and curation coverage. Future versions will progressively expand molecular features and strengthen patient-level annotations.
+
+---
+
+## Versioning & Updates
+
+- **Current version: v0.1**  
+  This release contains 354 curated entries, covering 16 cancer types and 58 unique HLA alleles (MHC-I). All entries include experimental immunogenicity evidence. Peptide lengths range from 8–13 amino acids, with 9-mers being the most common.  
+
+- **Update strategy**  
+  Future releases will expand coverage by adding new literature sources, additional mutation types, and molecular features not yet included in v0.1. Patient-level metadata will also be progressively enriched as more information becomes available.  
+
+- **Changelog**  
+  All modifications and additions will be documented in `CHANGELOG.md`, including new data sources, field extensions, and corrections.  
+
+- **Reproducibility**  
+  Historical versions of the dataset will remain available in the repository, allowing users to reproduce analyses and compare results across versions.  
+
+---
+
+## Community & Contributions
+
+NDD is an open community project, and we warmly welcome participation from researchers, clinicians, and data scientists. There are several ways to get involved:
+
+- **Feedback & suggestions**  
+  Since the dataset has been curated largely through manual annotation and standardization, omissions or errors may remain. We sincerely welcome any feedback or corrections.  
+  - Submit issues directly on GitHub Issues.  
+  - Or contact us by email at neoantigendd@gmail.com.  
+
+- **Data contributions**  
+  You are encouraged to share new data sources, patient cohorts, or predictive features that could enrich future releases.  
+
+- **Collaboration & discussion**  
+  We invite contributions in areas such as data curation, feature development, or method benchmarking. Open discussions and new ideas are welcome.  
+
+- **Pull requests (PRs)**  
+  If you wish to directly improve documentation, metadata, or scripts, you can submit a pull request — a standard GitHub workflow for proposing changes. This allows us to review your edits and merge them into the main project.  
+
+All contributions will be acknowledged in future version updates.  
+
+---
+
+## Citation
+
+If you use NDD in your research, please cite as:
+
+**Neoantigen Discovery Dataset (NDD), version 0.1**  
+GitHub Repository: [https://github.com/NeoDiscovery/NDD-v0.1](https://github.com/NeoDiscovery/NDD-v0.1)  
+Hugging Face Dataset: *to be released*  
+
+For formal publications, you may include the following BibTeX entry:
+
+```bibtex
+@misc{ndd2025,
+  title        = {Neoantigen Discovery Dataset (NDD), version 0.1},
+  author       = {{Neoantigen Discovery Dataset (NDD) Contributors}},
+  year         = {2025},
+  howpublished = {\url{https://github.com/NeoDiscovery/NDD-v0.1}},
+  note         = {Accessed: YYYY-MM-DD}
+}
+```
+
+---
+
+## License & Contact
+
+### Acknowledgments
+
+We thank the broader research community for making data and tools publicly available, which has greatly facilitated the development of NDD.  
+We also appreciate earlier efforts to compile and share neoantigen-related datasets, which provided valuable references during our work.  
+
+### License
+
+- Code and scripts: released under the MIT License.  
+- Dataset files (e.g., .tsv, .csv, .parquet): released under the Creative Commons Attribution 4.0 International License (CC BY 4.0).  
+  Users are free to share and adapt the dataset with proper attribution, in accordance with the license terms.  
+
+*Full license texts are available in the `LICENSE` and `DATA_LICENSE` files in this repository.*  
+
+### Contact
+
+For questions, feedback, or collaboration:  
+- GitHub Issues (preferred): open an issue  
+- Email: neoantigendd@gmail.com

data/patient_metadata.tsv

@@ -0,0 +1,165 @@
+patient_id	hla_reported_all	clinical_effects	
+D35NTCNS	HLA-A03:01	CR	
+6FHDB5EK	"HLA-B44:02,HLA-C06:02"	NA	
+DAXCBTFA	HLA-B44:02	NA	
+M4HZA63S	HLA-A02:01	NA	
+LFXUCEDT	HLA-A02:01	NA	
+V4WFIEVK	HLA-B52:01	NA	
+TSXDFNM6	HLA-A02:01	NA	
+I4HS6WN2	HLA-A02:01	PR	
+74LV6E74	HLA-A11:01	PR	
+J5PM7M2Y	HLA-A01:01	PR	
+U2DXRUJY	HLA-A03:01	CR	
+OZBCLZO4	HLA-A02:01	NA	
+VQQXFXNM	HLA-C14:03	NA	
+AMBE3DOP	HLA-B35:01	PR	
+V4UEFDFP	HLA-A02:01	CR	
+TSCSRBJX	"HLA-A02:01, HLA-A11:01, HLA-B18:01, HLA-B51:08, HLA-C07:01, HLA-C16:02"	NA	
+I7RVCSQA	HLA-A02:01	NA	
+OHQJPMZB	HLA-A02:01	NA	
+LE64IB2Z	HLA-A02:01	NA	
+ONOD4DCC	HLA-A02:01	NA	
+5K3HKC6D	HLA-A02:01	NA	
+Y3BFVLAR	HLA-A02:01	NA	
+J6KE3KIV	HLA-A01:01	NA	
+ASV5ARII	HLA-B08:01	NA	
+ABT5UFIF	"HLA-C05:01,HLA-C08:02"	NA	
+FFRFFTSB	HLA-A03:01	NA	
+Q5EFISOX	HLA-A02:01	NA	
+CUZXO5J4	HLA-B38:01	NA	
+ZEWNQSDP	HLA-B51:01	NA	
+6K4YGFNF	HLA-B07:02	CR	
+OGN36M6H	"HLA-A24:02,HLA-B38:01,HLA-C12:03"	CR	
+TOGZJCOT	"HLA-A01:01,HLA-A30:02,HLA-B15:01,HLA-C03:03,HLA-C05:01"	NR	
+DRAI56NY	HLA-B35:01	NA	
+JHK3ND2A	"HLA-A11:01,HLA-B07:02"	NA	
+WHG2YFVM	"HLA-A02:01,HLA-B15:01"	NA	
+7TRJS5PW	"HLA-B07:02,HLA-B35:01"	NA	
+A4433WE5	"HLA-A02:01,HLA-A29:02,HLA-B44:03"	NA	
+ZPRTYQED	HLA-A02:01	NA	
+U4R2OSJM	"HLA-A02:01,HLA-A03:01,HLA-B44:02,HLA-C14:02"	NA	
+PPKWDFGG	HLA-B15:01	NA	
+UCE5HKKE	HLA-B38:01	NA	
+5DJDBZPW	"HLA-A03:01,HLA-B27:05"	NA	
+PKPFBJDF	HLA-B07:02	NA	
+V4S75ZZC	"HLA-A31:01,HLA-C15:02"	NA	
+G3MJINSO	"HLA-A32:01,HLA-B35:01,HLA-B40:01"	NA	
+XYLTPHZM	"HLA-A24:02,HLA-B15:01"	NA	
+MILOVCNS	HLA-B15:01	NA	
+4RMAUA5K	HLA-A24:02	NA	
+RUNCUPXH	HLA-A24:02	NA	
+ATSSHWMT	HLA-A24:02	NA	
+P3V3F57E	HLA-A24:02	NA	
+SLHIGIK4	HLA-A24:02	NA	
+CAQZ3K3K	HLA-A24:02	NA	
+BBZ33E2J	HLA-A24:02	NA	
+VX46FFYO	HLA-A24:02	NA	
+KX4N5RSA	HLA-A24:02	NA	
+JJRPX753	HLA-B07:02	NA	
+QFNXKYKQ	HLA-A02:01	NA	
+PE5QXW4K	HLA-A02:01	NA	
+3RRUMLDF	"HLA-A24:02,HLA-B15:01"	NA	
+G4ZPYEDM	"HLA-A01:01,HLA-B56:01"	CR	
+G6DBM54G	HLA-B27:05	NA	
+F3XRZQLP	"HLA-B35:01,HLA-B41:02"	NA	
+URSBQAUG	HLA-B08:01	CR	
+DZL67UZM	"HLA-A02:01,HLA-B07:02,HLA-B44:02"	NA	
+NGE5BUYC	HLA-B07:02	NA	
+LC3W5VRX	HLA-A11:01	NA	
+LYXKV47S	HLA-A02:01	NA	
+N6T2LYFN	"HLA-A68:01,HLA-B37:01"	NA	
+OKQRI23E	HLA-A31:01	NA	
+VU5HBL4S	HLA-B39:06	NA	
+FRMBI3H3	"HLA-B57:01,HLA-A11:01"	NA	
+TUR5WXAA	HLA-A02:01	NA	
+FKQHULMK	HLA-B15:01	NA	
+FS2BPVBA	HLA-B15:01	NA	
+NBE6WCJN	"HLA-A02:01,HLA-A24:02"	NA	
+PC7T4K3B	HLA-A02:01	NA	
+DYEDTPM2	HLA-A02:01	NA	
+YQSS3IYL	HLA-A23:01	NA	
+VMHDDH4X	HLA-B07:02	NA	
+BFEZW33V	"HLA-B15:01,HLA-C03:03"	NA	
+TKJU56TT	HLA-A11:01	NA	
+Q2ITNIFI	HLA-A02:01	NA	
+TRB3HZII	HLA-A02:06	NA	
+GNQ75CJK	HLA-A01:01	NA	
+Q2CMUUIO	"HLA-A11:01,HLA-B27:05,HLA-C07:02"	NA	
+4DKUZH62	HLA-A02:11	NA	
+OP4YBPHD	HLA-A02:01	NA	
+BCIKPFC2	HLA-A02:01	NA	
+PFMINZ56	"HLA-B35:01,HLA-C04:01"	CR	
+DXUG6JS4	"HLA-A03:01,HLA-B51:01"	NA	
+65DW3BKE	HLA-A68:01	NA	
+6VPEE6FY	HLA-A02:01	NA	
+TE6XNJKV	HLA-A02:01	NA	
+RXNERV56	HLA-A02:01	NA	
+I4ZR7YUA	HLA-A68:01	NA	
+YG2NSK3U	HLA-B58:01	NA	
+TBH3PNXF	HLA-A03:01	NA	
+MYK5YWYV	HLA-A01:01	NA	
+ZLZLTBFT	HLA-A03:01	NA	
+KZYSWBUF	"HLA-A68:02,HLA-B53:01"	NA	
+E2RY6NVI	HLA-A02:01	NA	
+F3UNKL3Y	HLA-B15:01	NA	
+GVYOONV4	HLA-A02:01	NA	
+HVWWYCPV	"HLA-B14:02,HLA-C07:04"	NA	
+M4XLJUGS	HLA-A11:01	NA	
+DZZUH7EK	HLA-A30:01	NA	
+SFB6K5FA	HLA-A02:01	NA	
+SYIVONAA	HLA-A02:01	NA	
+EZJCG5SK	HLA-A11:01	NA	
+D4OVDCY4	HLA-A11:01	NA	
+47LXN6PR	HLA-A11:01	NA	
+GXEZJAFU	HLA-A02:01	NA	
+7KQJQFYF	HLA-A23:01	PR	
+IKJKM6CH	HLA-A02:01	CR	
+TFT6DO6A	HLA-A24:02	CR	
+R7NU4BAT	HLA-C06:02	NA	
+X3AIF5LA	HLA-A24:02	NA	
+XD6MQQ3M	"HLA-A25:01,HLA-A26:03,HLA-B35:01,HLA-C12:03"	NA	
+YE643ISX	HLA-B44:03	NA	
+7U4C6SW6	"HLA-A02:06,HLA-A02:24,HLA-A69:01"	NA	
+EXWFDAWT	"HLA-A30:01,HLA-A30:02"	NA	
+KX6Z6VLS	HLA-C07:02	NA	
+VTEIFCU7	HLA-A02:01	NA	
+D3NMWBFW	HLA-A03:01	NA	
+G33NW4LL	HLA-B35:01	NA	
+M2ABZXMV	"HLA-A24:02,HLA-B15:01"	NA	
+FNWFEVPE	"HLA-B52:01,HLA-C12:02"	NA	
+FU5SPZUA	"HLA-B18:01,HLA-C05:01,HLA-C12:03"	NA	
+6TP4IBLE	"HLA-A02:01,HLA-B08:01"	NA	
+OHRF7YEV	"HLA-A02:01,HLA-A11:01"	NA	
+FDCHYTN3	HLA-A68:02	NA	
+JAR22V53	HLA-A03:01	NA	
+3PCSYXUC	HLA-B44:03	NA	
+4V2JUJEQ	HLA-A11:01	NA	
+BRC35DEQ	HLA-A02:01	NA	
+7BEVRZEZ	HLA-A02:01	NA	
+3N22S6UG	HLA-B49:01	NA	
+YJOSS4J7	"HLA-A02:01,HLA-B07:02,HLA-B18:01,HLA-C07:02"	NA	
+OAF52DHL	"HLA-A02:01,HLA-A11:01,HLA-B07:02,HLA-B15:01,HLA-C07:02"	NA	
+IAXGKPRY	"HLA-A01:01,HLA-B44:02,HLA-C07:01"	NA	
+BOGIVUKN	"HLA-A03:01,HLA-B07:02"	NA	
+HRJPSEJN	"HLA-A24:02,HLA-C04:01"	NA	
+CWS5ZPHH	"HLA-B07:02,HLA-C07:02"	NA	
+5VPZ7ZNO	HLA-A02:03	NA	
+N6KFWJQT	HLA-A11:01	NA	
+4AYY5TKK	"HLA-A25:01,HLA-B51:01,HLA-C06:02,HLA-C15:02"	NA	
+EC43Q54U	HLA-C06:02	NA	
+BJGJ7IRT	HLA-C06:02	NA	
+B5DYJLWE	HLA-C15:02	NA	
+5UPLPDEU	HLA-A25:01	NA	
+GNOPTMJG	HLA-B51:01	NA	
+352WTYXA	HLA-A24:02	NA	
+GTFUJIR3	"HLA-A25:01,HLA-A68:01,HLA-B51:01"	NA	
+YJCPSI7I	"HLA-A01:01,HLA-B08:01"	NA	
+MP5A4MPY	HLA-A03:01	NA	
+JXPQ4HGM	"HLA-A03:01,HLA-B35:01"	NA	
+7LJZLSYK	HLA-A11:01	NA	
+GEXOATK7	HLA-A02:01	PR	
+ZWHDI5GT	HLA-B07:02	PR	
+XR365KHL	HLA-A24:02	NA	
+E2RCTTBH	HLA-A24:02	NA	
+VXLPQZMT	HLA-B40:01	NA	
+CNMRPXNS	HLA-A24:02	NA	

docs/field-schema.md

@@ -0,0 +1,26 @@
+# Appendix Table 1. Basic Fields
+
+| Field              | Category             | Description                                                                 |
+|--------------------|----------------------|-----------------------------------------------------------------------------|
+| pubmed_id          | data_source          | PubMed identifier of the reference                                          |
+| reference_name     | data_source          | Title or short name of the reference                                        |
+| patient_id         | sample_information   | Internal de-identified patient identifier                                   |
+| tumor_tissue       | sample_information   | Tumor sample tissue                                                         |
+| tumor_type_detail  | sample_information   | Original tumor type name reported in literature                             |
+| gene               | mutation_information | Gene symbol                                                                 |
+| mutation_type      | mutation_information | Mutation class                                                              |
+| mutation           | mutation_information | Mutation description (HGVS format or amino acid change, e.g., p.G12D)       |
+| position           | mutation_information | Mutation position within the peptide sequence (relative position in peptide)|
+| chromosome         | mutation_information | Chromosome number (1–22, X, Y)                                              |
+| genomic_coord      | mutation_information | Genomic coordinate                                                          |
+| ref                | mutation_information | Reference allele                                                            |
+| alt                | mutation_information | Alternative allele                                                          |
+| mt_peptide         | peptide_information  | Mutated peptide (8–14mer)                                                   |
+| wt_peptide         | peptide_information  | Corresponding wild-type peptide                                             |
+| length             | peptide_information  | Peptide length                                                              |
+| hla                | peptide_information  | Restricting HLA allele                                                      |
+| response_type      | peptide_information  | Response category                                                           |
+| assay_type         | assay_information    | Immunological assay type, harmonized from original literature into standardized categories. Multiple assay types may apply. |
+| effector_origin    | assay_information    | Source of effector T cells as described in the original study               |
+| stimulation_target | assay_information    | The target used to stimulate T cells, summarized from experimental description |
+| presentation_method| assay_information    | Experimental method used to demonstrate peptide presentation                |

docs/patient-metadata.md

@@ -0,0 +1,7 @@
+# Appendix Table 3. Patient-level Metadata
+
+| Field              | Description                                                                 |
+|--------------------|-----------------------------------------------------------------------------|
+| patient_id         | Unique patient identifier, corresponding to patient_id in Appendix Table 1  |
+| hla_reported_all   | Full set of HLA alleles reported in the literature                          |
+| clinical_effect    | Patient-level clinical outcome reported in the study (e.g., CR, PR, SD, PD; NA if not available) |

docs/predicted-features.md

@@ -0,0 +1,15 @@
+# Appendix Table 2. Predicted Features
+
+| Field                          | Definition                                                                                                                                  | Tool / Source                          |
+|--------------------------------|---------------------------------------------------------------------------------------------------------------------------------------------|----------------------------------------|
+| netmhcpan_rank                 | Percentile rank of peptide–MHC binding predicted by NetMHCpan                                                                               | NetMHCpan v4.1                         |
+| netmhcstabpan_stability        | Predicted stability of peptide–MHC complex                                                                                                  | NetMHCstabpan v1.0                     |
+| prime_rank                     | Predicted probability rank of TCR recognition                                                                                               | PRIME v1.0.1                           |
+| bigmhc_im_score                | Predicted immunogenicity probability of the peptide–MHC complex, as output by BigMHC (IM module). Continuous value between 0 and 1; higher scores indicate higher likelihood of immunogenicity. | BigMHC-IM (latest version, 2024 release) |
+| tap_score                      | Predicted TAP transport efficiency score                                                                                                    | NetCTLpan v1.1                         |
+| netchop_score                  | Predicted proteasomal cleavage score                                                                                                        | NetChop v3.1                           |
+| dai_netmhcpan                  | Differential agretopicity index (mutant vs wild-type, NetMHCpan). Calculation: log(rank(neo-pep)/rank(wt-pep)) if rank(neo-pep)>0 and rank(wt-pep)>0. 0 otherwise. Negative values indicate a large difference between neo-pep and wt-pep. | NetMHCpan v4.1                         |
+| anchor_mutation                | Whether mutation occurs at HLA anchor position                                                                                              | MixMHCpred v3.0                        |
+| eluted_ligand_match            | Whether peptide is observed in MHC ligand elution/MS experiments                                                                            | IEDB (MHC ligand elution, 2025 export) |
+| tcga_cancer_expression_tpm_median | Median log2(TPM+0.001) across TCGA primary tumor samples for the matching cancer type                                                    | TCGA (Data Release 43.0 – May 07, 2025) |
+| driver_status                  | Gene-level driver status annotation                                                                                                         | IntOGen 2024                           |