Datasets:

lordim
/

BenchRep-T

filename stringlengths 13 19	subject_id stringlengths 6 12	identifier stringlengths 11 17	participant_label stringlengths 6 12	disease stringclasses 2 values	fold int64 0 2	age float64 1.15 73.8 ⌀	sex stringclasses 2 values	ancestry stringclasses 4 values
P00001.tsv.gz	P00001	cmv__P00001	P00001	Healthy/Background	1	54.557153	M	null
P00002.tsv.gz	P00002	cmv__P00002	P00002	Healthy/Background	0	40.681725	M	null
P00003.tsv.gz	P00003	cmv__P00003	P00003	CMV	0	20.632444	M	Caucasian
P00004.tsv.gz	P00004	cmv__P00004	P00004	Healthy/Background	1	44.678987	F	null
P00005.tsv.gz	P00005	cmv__P00005	P00005	Healthy/Background	1	26.688569	M	Caucasian
P00006.tsv.gz	P00006	cmv__P00006	P00006	Healthy/Background	1	37.355236	F	null
P00007.tsv.gz	P00007	cmv__P00007	P00007	CMV	0	null	M	null
P00008.tsv.gz	P00008	cmv__P00008	P00008	Healthy/Background	0	31.917864	M	null
P00009.tsv.gz	P00009	cmv__P00009	P00009	Healthy/Background	1	44.323066	M	null
P00010.tsv.gz	P00010	cmv__P00010	P00010	Healthy/Background	0	null	F	null
P00011.tsv.gz	P00011	cmv__P00011	P00011	Healthy/Background	1	32.375086	F	null
P00012.tsv.gz	P00012	cmv__P00012	P00012	CMV	1	32.511978	F	null
P00013.tsv.gz	P00013	cmv__P00013	P00013	Healthy/Background	0	22.65024	F	null
P00014.tsv.gz	P00014	cmv__P00014	P00014	Healthy/Background	0	35.107461	M	Caucasian
P00015.tsv.gz	P00015	cmv__P00015	P00015	Healthy/Background	1	1.152635	F	Hispanic/Latino
P00016.tsv.gz	P00016	cmv__P00016	P00016	CMV	1	39.471595	M	Caucasian
P00017.tsv.gz	P00017	cmv__P00017	P00017	CMV	0	21.91102	M	null
P00018.tsv.gz	P00018	cmv__P00018	P00018	Healthy/Background	1	29.631759	M	Caucasian
P00019.tsv.gz	P00019	cmv__P00019	P00019	Healthy/Background	0	54.543463	M	null
P00020.tsv.gz	P00020	cmv__P00020	P00020	CMV	0	32.260096	M	Caucasian
P00021.tsv.gz	P00021	cmv__P00021	P00021	Healthy/Background	0	29.344285	M	null
P00022.tsv.gz	P00022	cmv__P00022	P00022	Healthy/Background	1	38.321697	F	null
P00023.tsv.gz	P00023	cmv__P00023	P00023	Healthy/Background	0	9.930185	F	Caucasian
P00024.tsv.gz	P00024	cmv__P00024	P00024	Healthy/Background	0	37.79603	M	Caucasian
P00025.tsv.gz	P00025	cmv__P00025	P00025	Healthy/Background	0	null	F	null
P00026.tsv.gz	P00026	cmv__P00026	P00026	CMV	1	25.859001	F	Hispanic/Latino
P00027.tsv.gz	P00027	cmv__P00027	P00027	CMV	1	33.664613	F	Caucasian
P00028.tsv.gz	P00028	cmv__P00028	P00028	Healthy/Background	1	null	F	null
P00029.tsv.gz	P00029	cmv__P00029	P00029	CMV	0	39.126626	M	null
P00030.tsv.gz	P00030	cmv__P00030	P00030	Healthy/Background	0	38.617385	F	null
P00031.tsv.gz	P00031	cmv__P00031	P00031	CMV	1	36.870637	M	null
P00032.tsv.gz	P00032	cmv__P00032	P00032	Healthy/Background	1	26.12731	M	null
P00033.tsv.gz	P00033	cmv__P00033	P00033	Healthy/Background	1	null	F	null
P00034.tsv.gz	P00034	cmv__P00034	P00034	CMV	1	null	F	null
P00035.tsv.gz	P00035	cmv__P00035	P00035	CMV	1	50.570842	F	null
P00036.tsv.gz	P00036	cmv__P00036	P00036	CMV	1	53.015743	M	Caucasian
P00037.tsv.gz	P00037	cmv__P00037	P00037	Healthy/Background	0	28.845996	M	null
P00038.tsv.gz	P00038	cmv__P00038	P00038	Healthy/Background	1	null	M	null
P00039.tsv.gz	P00039	cmv__P00039	P00039	Healthy/Background	0	null	F	null
P00040.tsv.gz	P00040	cmv__P00040	P00040	Healthy/Background	1	50.67488	M	null
P00041.tsv.gz	P00041	cmv__P00041	P00041	Healthy/Background	1	29.021218	F	null
P00042.tsv.gz	P00042	cmv__P00042	P00042	CMV	0	55.572895	F	Caucasian
P00043.tsv.gz	P00043	cmv__P00043	P00043	Healthy/Background	0	null	F	null
P00044.tsv.gz	P00044	cmv__P00044	P00044	Healthy/Background	1	null	M	null
P00045.tsv.gz	P00045	cmv__P00045	P00045	Healthy/Background	0	30.694045	M	null
P00046.tsv.gz	P00046	cmv__P00046	P00046	Healthy/Background	1	20.933607	M	null
P00047.tsv.gz	P00047	cmv__P00047	P00047	Healthy/Background	0	null	M	null
P00048.tsv.gz	P00048	cmv__P00048	P00048	Healthy/Background	1	34.847365	F	null
P00049.tsv.gz	P00049	cmv__P00049	P00049	CMV	0	null	M	null
P00050.tsv.gz	P00050	cmv__P00050	P00050	Healthy/Background	1	null	M	null
P00051.tsv.gz	P00051	cmv__P00051	P00051	Healthy/Background	0	40.40794	F	null
P00052.tsv.gz	P00052	cmv__P00052	P00052	Healthy/Background	0	null	M	null
P00053.tsv.gz	P00053	cmv__P00053	P00053	CMV	1	44.377823	F	null
P00054.tsv.gz	P00054	cmv__P00054	P00054	CMV	0	49.100616	M	null
P00055.tsv.gz	P00055	cmv__P00055	P00055	CMV	1	26.65024	M	null
P00056.tsv.gz	P00056	cmv__P00056	P00056	CMV	0	53.812457	M	null
P00057.tsv.gz	P00057	cmv__P00057	P00057	Healthy/Background	0	46.754278	M	null
P00058.tsv.gz	P00058	cmv__P00058	P00058	CMV	0	null	M	null
P00059.tsv.gz	P00059	cmv__P00059	P00059	Healthy/Background	1	null	M	null
P00060.tsv.gz	P00060	cmv__P00060	P00060	Healthy/Background	0	31.518138	M	null
P00062.tsv.gz	P00062	cmv__P00062	P00062	Healthy/Background	0	32.449008	M	null
P00063.tsv.gz	P00063	cmv__P00063	P00063	Healthy/Background	0	37.396304	F	null
P00064.tsv.gz	P00064	cmv__P00064	P00064	CMV	0	null	M	null
P00065.tsv.gz	P00065	cmv__P00065	P00065	Healthy/Background	0	42.8282	M	null
P00066.tsv.gz	P00066	cmv__P00066	P00066	Healthy/Background	1	44.952772	F	null
P00067.tsv.gz	P00067	cmv__P00067	P00067	Healthy/Background	1	24.884326	F	null
P00068.tsv.gz	P00068	cmv__P00068	P00068	CMV	1	null	F	null
P00069.tsv.gz	P00069	cmv__P00069	P00069	CMV	0	29.98768	M	null
P00070.tsv.gz	P00070	cmv__P00070	P00070	Healthy/Background	0	null	F	null
P00071.tsv.gz	P00071	cmv__P00071	P00071	CMV	0	null	M	null
P00072.tsv.gz	P00072	cmv__P00072	P00072	Healthy/Background	0	null	M	null
P00073.tsv.gz	P00073	cmv__P00073	P00073	Healthy/Background	1	null	F	null
P00074.tsv.gz	P00074	cmv__P00074	P00074	Healthy/Background	0	47.731691	M	null
P00075.tsv.gz	P00075	cmv__P00075	P00075	Healthy/Background	1	27.104723	M	null
P00076.tsv.gz	P00076	cmv__P00076	P00076	CMV	0	47.843943	F	null
P00077.tsv.gz	P00077	cmv__P00077	P00077	Healthy/Background	0	null	F	null
P00078.tsv.gz	P00078	cmv__P00078	P00078	Healthy/Background	0	35.29911	M	null
P00079.tsv.gz	P00079	cmv__P00079	P00079	Healthy/Background	0	null	M	null
P00080.tsv.gz	P00080	cmv__P00080	P00080	Healthy/Background	0	null	F	null
P00081.tsv.gz	P00081	cmv__P00081	P00081	Healthy/Background	0	27.457906	M	null
P00083.tsv.gz	P00083	cmv__P00083	P00083	CMV	1	32.629706	F	null
P00084.tsv.gz	P00084	cmv__P00084	P00084	CMV	0	null	M	null
P00085.tsv.gz	P00085	cmv__P00085	P00085	CMV	1	null	M	null
P00086.tsv.gz	P00086	cmv__P00086	P00086	Healthy/Background	1	null	M	null
P00087.tsv.gz	P00087	cmv__P00087	P00087	Healthy/Background	1	45.894593	M	null
P00088.tsv.gz	P00088	cmv__P00088	P00088	CMV	1	null	F	null
P00089.tsv.gz	P00089	cmv__P00089	P00089	CMV	1	51.367556	M	null
P00090.tsv.gz	P00090	cmv__P00090	P00090	CMV	0	27.154004	M	null
P00091.tsv.gz	P00091	cmv__P00091	P00091	Healthy/Background	1	14.297057	M	Caucasian
P00092.tsv.gz	P00092	cmv__P00092	P00092	Healthy/Background	0	9.785079	M	Caucasian
P00093.tsv.gz	P00093	cmv__P00093	P00093	Healthy/Background	0	6.324435	F	Caucasian
P00094.tsv.gz	P00094	cmv__P00094	P00094	CMV	1	12.542094	F	Caucasian
P00095.tsv.gz	P00095	cmv__P00095	P00095	CMV	0	54.236824	F	Caucasian
P00096.tsv.gz	P00096	cmv__P00096	P00096	CMV	0	null	M	null
P00097.tsv.gz	P00097	cmv__P00097	P00097	Healthy/Background	1	24.11499	F	null
P00098.tsv.gz	P00098	cmv__P00098	P00098	Healthy/Background	1	44.041068	M	null
P00099.tsv.gz	P00099	cmv__P00099	P00099	Healthy/Background	1	null	M	null
P00100.tsv.gz	P00100	cmv__P00100	P00100	Healthy/Background	0	null	F	null
P00101.tsv.gz	P00101	cmv__P00101	P00101	CMV	1	10.67488	F	Hispanic/Latino
P00102.tsv.gz	P00102	cmv__P00102	P00102	Healthy/Background	1	null	M	null

End of preview. Expand in Data Studio

BenchRep-T

BenchRep-T is a benchmark for TCRβ (T-cell receptor β-chain) immune-repertoire classification. It bundles the multi-disease Mal-ID cohort together with five external disease cohorts (sequenced on the Adaptive Biotechnologies immunoSEQ platform), providing per-specimen repertoire files, harmonized sample metadata, and auxiliary files used for the sequencing-depth scaling-law and driver-sequence experiments.

Each specimen (one sequenced repertoire) is one example; the classification target is the specimen-level disease label.

Dataset structure

BenchRep-T/
├── Mal-ID/
│   ├── metadata.tsv                             # 550 specimen annotations + labels
│   ├── repertoires/                             # 550 per-specimen TCRβ rearrangement tables (.tsv.gz)
│   ├── scaling_exp_depth_indices_max75k.json.gz # sampling indices for the depth scaling-law experiment
│   └── vdjdb_minervina_driver_seq_matches.csv   # repertoire ↔ public/driver-clone matches (driver-seq experiment)
├── immunoSEQ/
│   ├── Savola_RA/         # Rheumatoid Arthritis
│   │   ├── metadata.tsv
│   │   └── repertoires/   # 91 .tsv.gz
│   ├── Musvosvi_TB/       # Tuberculosis (progression)
│   │   ├── metadata.tsv
│   │   └── repertoires/   # 140 .tsv.gz
│   ├── Rawat_T1D/         # Type 1 Diabetes
│   │   ├── metadata.tsv
│   │   └── repertoires/   # 614 .tsv.gz
│   ├── Mitchell_T1D/      # Type 1 Diabetes
│   │   ├── metadata.tsv
│   │   └── repertoires/   # 196 .tsv.gz
│   └── Emerson_CMV/       # Cytomegalovirus exposure
│       ├── metadata.tsv
│       └── repertoires/   # 761 .tsv.gz
└── immuneML/              # immuneML-ready AIRR export of every cohort (see immuneML/README.md)
    ├── Mal-ID/, immunoSEQ/<cohort>/   # metadata.csv + AIRR repertoires/*.tsv.gz
    ├── tasks/<task>/                  # task metadata + fixed fold{0,1,2}_{train,test}.csv
    ├── specs/<task>.yaml              # runnable immuneML specifications
    └── README.md, MANIFEST.tsv, export_report.json

Cohorts

Group	Path	Classification task	Repertoires	Label composition	Repertoire format	Source study
Mal-ID	`Mal-ID/`	Multi-disease vs. Healthy/Background	550	Healthy/Background 197, HIV 98, Lupus 64, Covid19 58, Influenza 37, T1D 96	AIRR-style rearrangement table	Zaslavsky et al. 2025
Savola_RA	`immunoSEQ/Savola_RA/`	Rheumatoid Arthritis vs. Healthy	91	RA 71, Healthy 20	immunoSEQ export	Savola et al. 2017
Musvosvi_TB	`immunoSEQ/Musvosvi_TB/`	TB Progressor vs. Controller	140	Progressor 63, Controller 77	immunoSEQ export	Musvosvi et al. 2023
Rawat_T1D	`immunoSEQ/Rawat_T1D/`	Type 1 Diabetes vs. Control	614	T1D 426, Healthy/Background 188	immunoSEQ export (reduced)	Rawat et al. 2026
Mitchell_T1D	`immunoSEQ/Mitchell_T1D/`	Type 1 Diabetes vs. Healthy	196	T1D 171, Healthy/Background 25	immunoSEQ export	Mitchell et al. 2022
Emerson_CMV	`immunoSEQ/Emerson_CMV/`	CMV+ vs. Healthy/Background	761	CMV 340, Healthy/Background 421	immunoSEQ export	Emerson et al. 2017

Files

repertoires/*.tsv.gz — one gzip-compressed, tab-separated file per specimen; each row is one TCRβ sequencing read (Mal-ID; identical reads are not collapsed) or one unique rearrangement with its count (immunoSEQ cohorts). The file stem is the specimen identifier (for the immunoSEQ cohorts it matches the specimen_label/sample_name/filename column in that cohort's metadata.tsv; Rawat_T1D file names carry a _TCRB suffix).
metadata.tsv — one row per specimen with the classification label and per-cohort annotations (schemas differ across cohorts — see below).
Mal-ID/scaling_exp_depth_indices_max75k.json.gz — precomputed read-sampling indices (up to 75k reads/specimen) defining the subsampled repertoires used in the sequencing-depth scaling-law experiment.
Mal-ID/vdjdb_minervina_driver_seq_matches.csv — matches between repertoire clones and known public / antigen-specific "driver" TCRs (VDJdb + Minervina et al.), used for the driver-sequence experiment. Columns: disease, sample_cdr3, sample_vgene, sample_jgene, public_clone_cdr3, public_clone_vgene, public_clone_jgene, similarity, score, filename.

Repertoire schemas

Mal-ID (AIRR-style, ~120 columns) — key fields: sequence_id, repertoire_id, locus, v_call, d_call, j_call, cdr3, cdr3_aa, junction, junction_aa, productive, participant_label, specimen_time_point, plus full nucleotide alignment / insertion-deletion columns.

immunoSEQ cohorts (Savola_RA, Musvosvi_TB, Mitchell_T1D; Adaptive immunoSEQ export, ~54 columns) — key fields: nucleotide, cdr3_aa, count (templates/reads), frequencyCount (%), v_call, j_call, vGeneName, jGeneName, sequenceStatus, estimatedNumberGenomes, sequence, num_reads, repertoire_id, participant_label.

Rawat_T1D (reduced immunoSEQ-derived, 17 columns): cdr3_aa, count (templates/reads), frequency, nucleotide, v_call, d_gene, j_call, sequenceStatus, v_family, d_family, j_family, v_resolved, d_resolved, j_resolved, sequence, num_reads.

Emerson_CMV (minimal, 7 columns): cdr3_aa, v_call, j_call, sequence, num_reads, repertoire_id, participant_label.

Metadata schemas

All metadata.tsv files share participant_label, specimen_label, and disease (the classification target). Cross-validation fold assignments are provided per cohort (CV_fold / fold / malid_cross_validation_fold_id_when_in_test_set). Additional per-cohort columns:

Mal-ID / Mitchell_T1D: specimen_time_point, study_name, available_gene_loci, disease_subtype, age, sex, ancestry.
Savola_RA / Musvosvi_TB: immunoSEQ sample statistics (total_templates, productive_templates, fraction_productive, productive_simpson_clonality, sample_tags, sku, test_name, …).
Rawat_T1D: subject clinical/HLA fields (diabetes_status, sex, age, duration, HLA A/B/C/DPB1/DQB1/DRB1…, hla_high_risk_type, autoantibody statuses, ML_class).
Emerson_CMV: same base fields as Mal-ID/Mitchell_T1D (specimen_time_point, study_name, available_gene_loci, disease_subtype, age, sex, ancestry) plus repertoire_file, emerson_subject_id, cohort, cohort_name, race, ethnicity, race_and_ethnicity, known_cmv_status, metadata_source.

Usage

from huggingface_hub import snapshot_download
import pandas as pd, glob, os

# Download one cohort (metadata + repertoires)
local = snapshot_download(
    repo_id="lordim/BenchRep-T",
    repo_type="dataset",
    allow_patterns="immunoSEQ/Savola_RA/*",
)

meta = pd.read_csv(f"{local}/immunoSEQ/Savola_RA/metadata.tsv", sep="\t")
rep_files = glob.glob(f"{local}/immunoSEQ/Savola_RA/repertoires/*.tsv.gz")
rep = pd.read_csv(rep_files[0], sep="\t")   # pandas reads .gz transparently

immuneML-ready AIRR export

immuneML/ holds every cohort above in one AIRR Rearrangement schema that immuneML can import directly:

Repertoires: the same specimens and rows as the native files, in 20 AIRR columns that include all 14 required fields. cdr3_aa is unchanged. junction/junction_aa are rebuilt from the nucleotide data where they translate back to cdr3_aa: 99.78% of Mal-ID rows, and more than 99.99% in every other cohort.
Metadata: one comma-separated metadata.csv per cohort, with subject_id set to the specimen and BenchRep-T's fixed fold.
Tasks and specs: for each of the 11 benchmark tasks, the task's specimens, the three fixed train/test fold splits, and a runnable immuneML specification.

Tested with immuneML 3.0.27; install it with pandas<3. immuneML/README.md documents every conversion rule and caveat.

from huggingface_hub import snapshot_download

snapshot_download(
    repo_id="lordim/BenchRep-T", repo_type="dataset", local_dir="BenchRep-T",
    allow_patterns=["immuneML/immunoSEQ/Savola_RA/*", "immuneML/tasks/ra/*",
                    "immuneML/specs/ra.yaml"],
)
# then, inside BenchRep-T/:  immune-ml immuneML/specs/ra.yaml results/ra

License

Released under the MIT License.

References

Source studies for each cohort:

Mal-ID — Zaslavsky, Maxim E., et al. "Disease diagnostics using machine learning of B cell and T cell receptor sequences." Science 387.6736 (2025): eadp2407.
Savola_RA — Savola, Paula, et al. "Somatic mutations in clonally expanded cytotoxic T lymphocytes in patients with newly diagnosed rheumatoid arthritis." Nature Communications 8.1 (2017): 15869.
Musvosvi_TB — Musvosvi, Munyaradzi, et al. "T cell receptor repertoires associated with control and disease progression following Mycobacterium tuberculosis infection." Nature Medicine 29.1 (2023): 258-269.
Rawat_T1D — Rawat, Puneet, et al. "Identification of a type 1 diabetes–associated T cell receptor repertoire signature from the human peripheral blood." Science Advances 12.7 (2026): eadx7448.
Mitchell_T1D — Mitchell, Angela M., et al. "Temporal development of T cell receptor repertoires during childhood in health and disease." JCI Insight 7.18 (2022): e161885.
Emerson_CMV — Emerson, Ryan O., et al. "Immunosequencing identifies signatures of cytomegalovirus exposure history and HLA-mediated effects on the T cell repertoire." Nature Genetics 49.5 (2017): 659-665.

Citation

BenchRep-T is described in the following paper, to appear in the NeurIPS 2026 Evaluations and Datasets Track:

Alejandro Buendia*, Chiho Im*, Anshul Kundaje, Scott D Boyd, and Liel Cohen-Lavi. "BenchRep-T: A Systematic Evaluation of T-Cell Repertoire-Based Disease Diagnostics." Advances in Neural Information Processing Systems (NeurIPS), Evaluations and Datasets Track, 2026.

* Equal contribution (co-first authors).

@inproceedings{buendia2026benchrep,
  title     = {BenchRep-T: A Systematic Evaluation of T-Cell Repertoire-Based Disease Diagnostics},
  author    = {Buendia, Alejandro and Im, Chiho and Kundaje, Anshul and Boyd, Scott D and Cohen-Lavi, Liel},
  booktitle = {Advances in Neural Information Processing Systems, Evaluations and Datasets Track},
  year      = {2026}
}

Code: github.com/boyd-lab/BenchRep-T. A preprint is available on bioRxiv.

Downloads last month: 85