isalgo/tcren_structures

TCR:peptide:MHC structure sets and benchmarks for TCRen2 (structure-based prediction of TCR recognition). Fetch with tcren / the manuscript scripts/bootstrap_data.py.

Contents rule: compressed structures (LFS), descriptions, and flat TSV tables for identities, provenance, generator confidence and filename conversions. Assay readouts for TCRvdb / IMMREP26 remain private. No notebooks, figures or computed benchmark analysis tables.

Layout

folder	task	contents
Native2026/, Canonical2026/	derivation / ergodicity	non-redundant TCR:pMHC structures (.gz)
Native2022/, PolyV2022/	2022-paper reproduction	structure sets (.gz)
tcrvdb/	current receptor benchmark	2,636 structures / eight epitopes; identity, provenance and TCRmodel2 confidences only; see its README
tcrvdb_v1/	old receptor modelling run	618 old filename aliases / 30 additional coordinates; conversion table; do not pool with the current run
vdjdb_binder_benchmark/	TCR-ranking (real binders vs mock)	1,089 PDBs (.tar.gz) + metadata.tsv (labels, ipTM, tcren scores, covered/uncovered flag); see its README.md
vdjdb_mut2023_benchmark/	TCR-recognition control, one epitope	81 PDBs (.tar.gz) + metadata.tsv: HPVGEADYFEY crystals, their Rosetta mutants and TCRmodel2 mock decoys; see its README.md
immrep23_negatives/	TCR-ranking, extra decoys	832 of 2,985 modelled IMMREP mismatched pairings (run in progress); lifts the per-epitope decoy ceiling
cpl/	peptide-ranking (epitope)	archived peptide models; see cpl/content.md
as_case/	MHC-ranking (B*27:05 vs :02)	as_case.tar.gz + content.md
Bobisse/, Bigot/	neoantigen	cohort structures (.gz)

Uses

These collections support ranking receptors for a pMHC, peptides for a receptor, and MHC alleles for a receptor–peptide pair. Modelling runs and dataset cohorts retain their own identities and provenance; overlapping runs must not be pooled as independent observations. Collection READMEs define the supplied inputs.