{"success":true,"data":{"author":"Oxford Protein Informatics Group (OAS); Parquet conversion by Converge Bio","bucketAccessType":"public","bucketName":"ConvergeBio/oas-unpaired","bucketPrefix":"","category":"Biology","cloudProvider":"huggingface","createdAt":"2026-11-16T14:13:15.565Z","description":"The unpaired portion of the Observed Antibody Space (OAS), the Oxford Protein Informatics Group's database of cleaned, annotated antibody repertoire sequencing data, converted to Parquet by Converge Bio. It holds about 2.07 billion heavy-chain and 357 million light-chain sequences, about 2.7 TB in total. Each row carries the AIRR-style annotation OAS provides, including the aligned amino-acid sequence, CDR3, V and J gene calls, and study and subject metadata. The repo exposes configs for all heavy chains, all light chains, and one config per source study (for example Briney et al., 2019), so a single study or a column subset can be read without touching the rest. OAS is the standard pretraining corpus for antibody language models. OAS data is CC BY 4.0; cite Olsen, Boyles and Deane, Protein Science 2022.","displayTitle":"Observed Antibody Space (OAS) unpaired heavy and light chains, Parquet","huggingFaceUrl":"https://huggingface.co/datasets/ConvergeBio/oas-unpaired","id":"73b16e27-5510-5673-912b-fb2c44aea3ef","license":"CC-BY-4.0","licenseEvidenceUrl":"https://opig.stats.ox.ac.uk/webapps/oas/","objectCountEstimate":4032,"probeStatus":"ok","referenceUrl":"https://doi.org/10.1002/pro.4205","region":"us","schemaCard":{"accessHints":[{"language":"python-datasets","snippet":"from datasets import load_dataset\nds = load_dataset(\"ConvergeBio/oas-unpaired\")"},{"language":"python-hub","snippet":"from huggingface_hub import snapshot_download\nsnapshot_download(repo_id=\"ConvergeBio/oas-unpaired\", repo_type=\"dataset\")"}],"caveats":["HuggingFace metadata: author=ConvergeBio, license:cc-by-4.0, size_categories:1B<n<10B, format:parquet, downloads=6823"],"format":"mixed","generatedAt":"2026-11-16T14:13:18.413Z","layout":{"example":"data/unpaired_heavy/Banerjee et al., 2017/train-00000.parquet","pattern":"{repoId}/{filename}"},"mountPath":"/mnt/archive","sampleHeaders":[{"bytes":"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","encoding":"hex","key":"data/unpaired_heavy/Banerjee et al., 2017/train-00000.parquet","truncatedAt":512}],"sampleObjects":["data/unpaired_heavy/Banerjee et al., 2017/train-00000.parquet","data/unpaired_heavy/Banerjee et al., 2017/train-00001.parquet","data/unpaired_heavy/Banerjee et al., 2017/train-00002.parquet","data/unpaired_heavy/Banerjee et al., 2017/train-00003.parquet","data/unpaired_heavy/Banerjee et al., 2017/train-00004.parquet"],"sidecars":[{"bytes":"---\nlicense: cc-by-4.0\nlanguage:\n- en\ntags:\n- antibody\n- immunology\n- bcr\n- airr\n- oas\n- proteomics\n- proteins\n- protein\nsize_categories:\n- 1B<n<10B\nconfigs:\n- config_name: default\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/**/*.parquet\n  - split: light\n    path: data/unpaired_light/**/*.parquet\n- config_name: heavy\n  data_files:\n  - split: train\n    path: data/unpaired_heavy/**/*.parquet\n- config_name: light\n  data_files:\n  - split: train\n    path: data/unpaired_light/**/*.parquet\n- config_name: Banerjee et al., 2017\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Banerjee et al., 2017/*.parquet\n- config_name: Bashford et al., 2013\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Bashford et al., 2013/*.parquet\n- config_name: Bender et al., 2020\n  data_files:\n  - split: light\n    path: data/unpaired_light/Bender et al., 2020/*.parquet\n- config_name: Bernardes et al., 2020\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Bernardes et al., 2020/*.parquet\n  - split: light\n    path: data/unpaired_light/Bernardes et al., 2020/*.parquet\n- config_name: Bernat et al., 2019\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Bernat et al., 2019/*.parquet\n  - split: light\n    path: data/unpaired_light/Bernat et al., 2019/*.parquet\n- config_name: Bhiman et al., 2015\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Bhiman et al., 2015/*.parquet\n  - split: light\n    path: data/unpaired_light/Bhiman et al., 2015/*.parquet\n- config_name: Bolland et al., 2016\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Bolland et al., 2016/*.parquet\n- config_name: Bonsignori et al., 2016\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Bonsignori et al., 2016/*.parquet\n- config_name: Briney et al., 2019\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Briney et al., 2019/*.parquet\n  - split: light\n    path: data/unpaired_light/Briney et al., 2019/*.parquet\n- config_name: Buchheim et al., 2020\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Buchheim et al., 2020/*.parquet\n- config_name: Chen et al., 2020\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Chen et al., 2020/*.parquet\n  - split: light\n    path: data/unpaired_light/Chen et al., 2020/*.parquet\n- config_name: Collins et al., 2015\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Collins et al., 2015/*.parquet\n- config_name: Corcoran et al., 2016\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Corcoran et al., 2016/*.parquet\n  - split: light\n    path: data/unpaired_light/Corcoran et al., 2016/*.parquet\n- config_name: Cui et al., 2019\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Cui et al., 2019/*.parquet\n  - split: light\n    path: data/unpaired_light/Cui et al., 2019/*.parquet\n- config_name: Davis et al., 2019\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Davis et al., 2019/*.parquet\n- config_name: Doria-Rose et al., 2015\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Doria-Rose et al., 2015/*.parquet\n  - split: light\n    path: data/unpaired_light/Doria-Rose et al., 2015/*.parquet\n- config_name: Eccles et al., 2020\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Eccles et al., 2020/*.parquet\n  - split: light\n    path: data/unpaired_light/Eccles et al., 2020/*.parquet\n- config_name: Eliyahu et al., 2015\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Eliyahu et al., 2015/*.parquet\n- config_name: Ellebedy et al., 2016\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Ellebedy et al., 2016/*.parquet\n- config_name: Fisher et al., 2017\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Fisher et al., 2017/*.parquet\n  - split: light\n    path: data/unpaired_light/Fisher et al., 2017/*.parquet\n- config_name: Galson et al., 2015\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Galson et al., 2015/*.parquet\n- config_name: Galson et al., 2015a\n  data_files:\n  - split: heavy\n    path: data/unpaired_heavy/Galson et al., 2015a/*.parquet\n- confi","encoding":"utf8","key":"README.md","name":"README.md","size":42491,"truncated":true}],"sourceObjectKey":"data/unpaired_heavy/Banerjee et al., 2017/train-00000.parquet","starterCell":"from datasets import load_dataset\nds = load_dataset(\"ConvergeBio/oas-unpaired\")\nprint(ds)","starterSamplePath":"data/unpaired_heavy/Banerjee et al., 2017/train-00000.parquet","totalObjects":4032},"summary":"The unpaired portion of the Observed Antibody Space (OAS), the Oxford Protein Informatics Group's database of cleaned, annotated antibody repertoire sequencing data, converted to…","title":"Observed Antibody Space (OAS) unpaired heavy and light chains, Parquet","updatedAt":"2026-11-16T15:18:17.175Z","visibility":"public","websiteUrl":"https://opig.stats.ox.ac.uk/webapps/oas/"}}