OATML-Markslab/ProteinGym / README.md

What this source contributed

Records in this release that cite this source. A record can cite several sources.

• 27 models
• 27 configurations
• 1 benchmark
• 3 protocols
• 3 dataset subsets
• 230 evidence claims
• 1 use case

Claims

Statements this source supports, each with where in the source it was checked.

• AIDO Protein-RAG (16B) AMFR official snapshot: attributes.origin: Origin unreported (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• AIDO Protein-RAG (16B) AMFR official snapshot: description: Description Official ProteinGym assay-level Spearman for the AMFR assay at commit 144fe22; transcribed, not reproduced. (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• AIDO Protein-RAG: description: Description Retrieval-augmented protein language model from the AIDO.Protein-RAG work, which combines a pretrained protein language model with retrieved multiple sequence alignments; ProteinGym lists AIDO with alignment and structure inputs and scores it as a zero-shot substitution baseline. (uc20260930-source-proteingym-amfr-spearman: Benchmark CSV header (model scored in the zero-shot DMS substitution benchmark); run-doc-proteingym-readme-md-144fe22b: README 'Leaderboard' baseline table row AIDO (MSA & Structure); qa-descriptions-20261010-source-aido-protein-rag: abstract)
• Baseline: Predict missense variant effects in a protein of interest: Baseline Site-independent alignment model (run-doc-proteingym-readme-md-144fe22b: README lines 66-70)
• CARP (38M) AMFR official snapshot: attributes.origin: Origin unreported (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• CARP (38M) AMFR official snapshot: description: Description Official ProteinGym assay-level Spearman for the AMFR assay at commit 144fe22; transcribed, not reproduced. (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• CARP (76M) AMFR official snapshot: attributes.origin: Origin unreported (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• CARP (76M) AMFR official snapshot: description: Description Official ProteinGym assay-level Spearman for the AMFR assay at commit 144fe22; transcribed, not reproduced. (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• CARP (600K) AMFR official snapshot: attributes.origin: Origin unreported (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• CARP (600K) AMFR official snapshot: description: Description Official ProteinGym assay-level Spearman for the AMFR assay at commit 144fe22; transcribed, not reproduced. (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• CARP (640M) AMFR official snapshot: attributes.origin: Origin unreported (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• CARP (640M) AMFR official snapshot: description: Description Official ProteinGym assay-level Spearman for the AMFR assay at commit 144fe22; transcribed, not reproduced. (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• CARP: description: Description Convolutional protein language model pretrained by masked language modelling on protein sequences, scored in ProteinGym as a zero-shot single-sequence substitution baseline. (uc20260930-source-proteingym-amfr-spearman: Benchmark CSV header (model scored in the zero-shot DMS substitution benchmark); run-doc-proteingym-readme-md-144fe22b: README 'Leaderboard' baseline table (model name, input modalities, reference); qa-descriptions-20261010-source-carp: abstract)
• DeepSequence (ensemble) AMFR official snapshot: attributes.origin: Origin unreported (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• DeepSequence (ensemble) AMFR official snapshot: description: Description Official ProteinGym assay-level Spearman for the AMFR assay at commit 144fe22; transcribed, not reproduced. (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• DeepSequence (single) AMFR official snapshot: attributes.origin: Origin unreported (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• DeepSequence (single) AMFR official snapshot: description: Description Official ProteinGym assay-level Spearman for the AMFR assay at commit 144fe22; transcribed, not reproduced. (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• DeepSequence: description: Description Latent-variable deep generative model of a protein family's sequences, learned without labels from evolutionary sequence data, that predicts mutation effects; ProteinGym scores it as a zero-shot baseline from a multiple sequence alignment. (uc20260930-source-proteingym-amfr-spearman: Benchmark CSV header (model scored in the zero-shot DMS substitution benchmark); run-doc-proteingym-readme-md-144fe22b: README 'Leaderboard' baseline table (model name, input modalities, reference); qa-descriptions-20261010-source-deepsequence: abstract)
• ESCOTT AMFR official snapshot: attributes.origin: Origin unreported (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• ESCOTT AMFR official snapshot: description: Description Official ProteinGym assay-level Spearman for the AMFR assay at commit 144fe22; transcribed, not reproduced. (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• ESCOTT: description: Description Missense effect predictor that ProteinGym lists with multiple sequence alignment and structure inputs under the PRESCOTT reference, which describes an epistatic and structural model of missense effects that also uses population allele frequencies; scored as a zero-shot substitution baseline. (uc20260930-source-proteingym-amfr-spearman: Benchmark CSV header (model scored in the zero-shot DMS substitution benchmark); run-doc-proteingym-readme-md-144fe22b: README 'Leaderboard' baseline table (model name, input modalities, reference); qa-descriptions-20261010-source-escott: abstract)
• ESM-1b AMFR official snapshot: attributes.origin: Origin unreported (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• ESM-1b AMFR official snapshot: description: Description Official ProteinGym assay-level Spearman for the AMFR assay at commit 144fe22; transcribed, not reproduced. (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))
• ESM-1b: description: Description Deep contextual protein language model trained without labels on 250 million protein sequences, whose learned representations support mutational-effect prediction; ProteinGym scores it as a zero-shot single-sequence substitution baseline. (uc20260930-source-proteingym-amfr-spearman: Benchmark CSV header (model scored in the zero-shot DMS substitution benchmark); run-doc-proteingym-readme-md-144fe22b: README 'Leaderboard' baseline table (model name, input modalities, reference); qa-descriptions-20261010-source-esm-1b: abstract)
• ESM-1v (ensemble) AMFR official snapshot: attributes.origin: Origin unreported (DMS_substitutions_Spearman_DMS_level.csv: per-model columns with no scorer field; README.md at 144fe22, 'New baselines' (lines 159-170) and Acknowledgements (line 222))

Models, benchmarks and other records

Evidence

Source checking verifies the cited claim or transcription. It does not establish independent reproduction.