Configuration

ESM2

This FUJISAN-study baseline compares proteins by cosine similarity between mean ESM-2 embeddings.

Sources

Enhanced prediction of protein functional identity through the integration of sequence and structural features · Materials and methods/Prediction with ESM-2 (paragraph 2); Results and discussion/Limitations and perspectives (paragraph 1)

1 evaluation · 8 results

How it worksEvaluated procedure (conceptual)

Conceptual input–method–output guide. Check the procedure text and linked evaluation for fitted components, additional inputs and exact settings.

Sources

Enhanced prediction of protein functional identity through the integration of sequence and structural features · Materials and methods/Prediction with ESM-2 (paragraph 2); Materials and methods/Prediction with DeepFRI (paragraph 1)

Overview

Model type

Protein sequence transformer; this record is the paper-specific evaluated configuration.

Sources

facebookresearch/esm README.md · README.md model description

Biological inputs

Pairs of protein amino-acid sequences

Sources

Enhanced prediction of protein functional identity through the integration of sequence and structural features · Materials and methods/Prediction with ESM-2 (paragraph 1); Introduction (paragraph 4)

Outputs

Pairwise cosine similarity used as a same-function score

Sources

Enhanced prediction of protein functional identity through the integration of sequence and structural features · Materials and methods/Prediction with ESM-2 (paragraph 2); Materials and methods/Prediction with DeepFRI (paragraph 1)

Access

Official upstream implementation and usage documentation: https://github.com/facebookresearch/esm/blob/2b369911bb5b4b0dda914521b9475cad1656b2ac/README.md. This pinned documentation revision is not automatically the evaluated weight revision.

Sources

facebookresearch/esm README.md · README.md; installation, model download and usage instructions

limited source coverage · Automated source review, 2026-09-16. All specifications and missing details

Evaluations and results

1 evaluation · 8 results. Different protocols are not a single leaderboard.

Filter evaluations

Tested model or methodProtocolDatasetEvaluation setupMetricEvidence origin

Applied filters: All linked evaluations

Exact evaluated configurations and original reported results
Tested configuration	Protocol and dataset	Finding	Evidence and details
Configuration: ESM2	Protocol: Enzyme-pair functional identity: original held-out test (Enzyme functional identity prediction) Dataset: FUJISAN test sub-dataset	0.799 AUROC unitless · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source ESM2: Enzyme functional identity prediction LightGBM FUJISAN and comparison methods; predictions thresholded to maximizeF 1. Protein-pair splitting is not evidence of disjoint proteins or families. 41,600 protein pairs; balanced functional-identity classes; random 56.25/18.75/25% training/validation/test split. Aggregation: Not reported Enhanced prediction of protein functional identity through the integration of sequence and structural features; Enhanced prediction of protein functional identity through the integration of sequence and structural features · Table 1, ESM2 row, AUROC column
Configuration: ESM2	Protocol: Enzyme-pair functional identity: original held-out test (Enzyme functional identity prediction) Dataset: FUJISAN test sub-dataset	79.3% REC (%) percent · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source ESM2: Enzyme functional identity prediction LightGBM FUJISAN and comparison methods; predictions thresholded to maximizeF 1. Protein-pair splitting is not evidence of disjoint proteins or families. 41,600 protein pairs; balanced functional-identity classes; random 56.25/18.75/25% training/validation/test split. Aggregation: Not reported Enhanced prediction of protein functional identity through the integration of sequence and structural features · Table 1, row ESM2, column REC (%); XML row5 column4
Configuration: ESM2	Protocol: Enzyme-pair functional identity: original held-out test (Enzyme functional identity prediction) Dataset: FUJISAN test sub-dataset	0.815 AUPR dimensionless · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source ESM2: Enzyme functional identity prediction LightGBM FUJISAN and comparison methods; predictions thresholded to maximizeF 1. Protein-pair splitting is not evidence of disjoint proteins or families. 41,600 protein pairs; balanced functional-identity classes; random 56.25/18.75/25% training/validation/test split. Aggregation: Not reported Enhanced prediction of protein functional identity through the integration of sequence and structural features · Table 1, row ESM2, column AUPR; XML row5 column9
Configuration: ESM2	Protocol: Enzyme-pair functional identity: original held-out test (Enzyme functional identity prediction) Dataset: FUJISAN test sub-dataset	0.432 MCC dimensionless · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source ESM2: Enzyme functional identity prediction LightGBM FUJISAN and comparison methods; predictions thresholded to maximizeF 1. Protein-pair splitting is not evidence of disjoint proteins or families. 41,600 protein pairs; balanced functional-identity classes; random 56.25/18.75/25% training/validation/test split. Aggregation: Not reported Enhanced prediction of protein functional identity through the integration of sequence and structural features · Table 1, row ESM2, column MCC; XML row5 column7
Configuration: ESM2	Protocol: Enzyme-pair functional identity: original held-out test (Enzyme functional identity prediction) Dataset: FUJISAN test sub-dataset	68.4% PRE (%) percent · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source ESM2: Enzyme functional identity prediction LightGBM FUJISAN and comparison methods; predictions thresholded to maximizeF 1. Protein-pair splitting is not evidence of disjoint proteins or families. 41,600 protein pairs; balanced functional-identity classes; random 56.25/18.75/25% training/validation/test split. Aggregation: Not reported Enhanced prediction of protein functional identity through the integration of sequence and structural features · Table 1, row ESM2, column PRE (%); XML row5 column3
Configuration: ESM2	Protocol: Enzyme-pair functional identity: original held-out test (Enzyme functional identity prediction) Dataset: FUJISAN test sub-dataset	0.735 F1 dimensionless · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source ESM2: Enzyme functional identity prediction LightGBM FUJISAN and comparison methods; predictions thresholded to maximizeF 1. Protein-pair splitting is not evidence of disjoint proteins or families. 41,600 protein pairs; balanced functional-identity classes; random 56.25/18.75/25% training/validation/test split. Aggregation: Not reported Enhanced prediction of protein functional identity through the integration of sequence and structural features · Table 1, row ESM2, column F1; XML row5 column6
Configuration: ESM2	Protocol: Enzyme-pair functional identity: original held-out test (Enzyme functional identity prediction) Dataset: FUJISAN test sub-dataset	71.3% ACC (%) percent · higher Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source ESM2: Enzyme functional identity prediction LightGBM FUJISAN and comparison methods; predictions thresholded to maximizeF 1. Protein-pair splitting is not evidence of disjoint proteins or families. 41,600 protein pairs; balanced functional-identity classes; random 56.25/18.75/25% training/validation/test split. Aggregation: Not reported Enhanced prediction of protein functional identity through the integration of sequence and structural features · Table 1, row ESM2, column ACC (%); XML row5 column2
Configuration: ESM2	Protocol: Enzyme-pair functional identity: original held-out test (Enzyme functional identity prediction) Dataset: FUJISAN test sub-dataset	36.7% FPR (%) percent · lower Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source ESM2: Enzyme functional identity prediction LightGBM FUJISAN and comparison methods; predictions thresholded to maximizeF 1. Protein-pair splitting is not evidence of disjoint proteins or families. 41,600 protein pairs; balanced functional-identity classes; random 56.25/18.75/25% training/validation/test split. Aggregation: Not reported Enhanced prediction of protein functional identity through the integration of sequence and structural features · Table 1, row ESM2, column FPR (%); XML row5 column5

Source checking is not independent reproduction. Release 2026-09-29-06401fd5b220.

Use this model

How it works, versions and access

Underlying model: ESM-2. Results on this page belong to this configuration and its evaluated settings.

How it works

How the evaluated method works

esm2_t33_650M_UR50D emits 1,280-dimensional residue vectors. Mean pooling gives a protein vector; cosine similarity estimates whether two proteins share an enzymatic function.

Sources

Enhanced prediction of protein functional identity through the integration of sequence and structural features · Materials and methods/Prediction with ESM-2 (paragraph 2); Materials and methods/Prediction with DeepFRI (paragraph 1)

Underlying method and version boundaries

ESM-2 is a transformer protein language-model family. The official repository exposes residue embeddings, sequence-level pooling and models at several sizes; the study configuration determines which of these is evaluated.

Sources

facebookresearch/esm README.md · README.md; introduction, model description, pretrained-model and usage sections at pinned revision

What was evaluated

The linked evaluation record identifies ESM2: Enzyme functional identity prediction. Its dataset, split, adaptation and evidence origin remain attached to the reported results.

Sources

Enhanced prediction of protein functional identity through the integration of sequence and structural features · The named evaluation’s methods and comparison table; exact preserved evaluation IDs: evaluation-lit-020

Strengths, limitations and unresolved questions

Strengths and limitations

Strengths and considerations

A simple sequence-only representation baseline with an explicit pooling and scoring rule.
Sources
Enhanced prediction of protein functional identity through the integration of sequence and structural features · Materials and methods/Prediction with ESM-2 (paragraph 1); Materials and methods/Feature engineering/Pocket similarity features (paragraph 1)

Limitations and conditions

Cosine similarity is a proxy for functional similarity, not a calibrated catalytic-function probability.
Sources
Enhanced prediction of protein functional identity through the integration of sequence and structural features · Materials and methods/Prediction with ESM-2 (paragraph 2); Conclusions (paragraph 1)

Profile review details

Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction.

Stable record: reported-model-ccd1160ad4ec27

Specifications

Inputs, training, access and other details

Explanatory profile: limited source coverage · Automated source review, 2026-09-16. Review applies to the cited claims; unresolved fields are listed below. Numerical results retain their own review status.

Inputs, outputs and configuration
Property	Description and evidence
Model type	Protein sequence transformer; this record is the paper-specific evaluated configuration. Sources facebookresearch/esm README.md · README.md model description
Architecture / procedure	esm2_t33_650M_UR50D emits 1,280-dimensional residue vectors. Mean pooling gives a protein vector; cosine similarity estimates whether two proteins share an enzymatic function. Sources Enhanced prediction of protein functional identity through the integration of sequence and structural features · Materials and methods/Prediction with ESM-2 (paragraph 2); Materials and methods/Prediction with DeepFRI (paragraph 1)
Biological inputs	Pairs of protein amino-acid sequences Sources Enhanced prediction of protein functional identity through the integration of sequence and structural features · Materials and methods/Prediction with ESM-2 (paragraph 1); Introduction (paragraph 4)
Outputs	Pairwise cosine similarity used as a same-function score Sources Enhanced prediction of protein functional identity through the integration of sequence and structural features · Materials and methods/Prediction with ESM-2 (paragraph 2); Materials and methods/Prediction with DeepFRI (paragraph 1)
Parameters	650 million backbone parameters; no FUJISAN LightGBM features are part of this comparator. Sources Enhanced prediction of protein functional identity through the integration of sequence and structural features · Results and discussion/Analysis of feature importance (paragraph 1); Materials and methods/Model training and hyperparameter optimization (paragraph 1)
Known versions / configuration	esm2_t33_650M_UR50D · Not reported in inspected sources Sources Enhanced prediction of protein functional identity through the integration of sequence and structural features · Model identification in the comparison table and corresponding Methods; immutable checkpoint revision is not supplied by the table label.
Training data / fitting	The comparator uses pretrained ESM-2 sequence embeddings and cosine similarity; it does not fit FUJISAN’s LightGBM head or a task-specific interaction classifier. Sources Enhanced prediction of protein functional identity through the integration of sequence and structural features · ESM2 embedding comparator and cosine-similarity procedure
Context limits	A maximum input/context length for this exact evaluated configuration is not established by the inspected sources. · Not reported in inspected sources Sources (2) Enhanced prediction of protein functional identity through the integration of sequence and structural features; facebookresearch/esm README.md · Materials and methods/Dataset construction; Materials and methods/Feature engineering/Full-length sequence similarity features; Materials and methods/Feature engineering/Domain structural similarity features; Materials and methods/Feature engineering/Pocket similarity features; Materials and methods/Model training and hyperparameter optimization; Materials and methods/Performance assessment; Materials and methods/Prediction with DeepFRI; Materials and methods/Prediction with ESM-2; inspected for explicit maximum input length (dataset lengths and family-wide limits are not substituted); README.md at pinned repository revision
Access	Official upstream implementation and usage documentation: https://github.com/facebookresearch/esm/blob/2b369911bb5b4b0dda914521b9475cad1656b2ac/README.md. This pinned documentation revision is not automatically the evaluated weight revision. Sources facebookresearch/esm README.md · README.md; installation, model download and usage instructions
Code licence	MIT (upstream repository code at the cited revision; this does not establish every dependency or historical checkpoint licence). Sources facebookresearch/esm LICENSE · LICENSE; complete licence text
Weights licence	The inspected model-access documentation does not explicitly identify terms for this exact evaluated checkpoint or fitted head; repository code terms are shown separately. · Not reported in inspected sources Sources facebookresearch/esm README.md · README.md; checkpoint/access documentation and licence scope

Evidence

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

Evidence table

Inspect claims, sources and review details

Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.

One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.

22 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-09-29-06401fd5b220
Property and statement	Original source and location	Review and provenance
Diagram caption Conceptual input–method–output guide. Check the procedure text and linked evaluation for fitted components, additional inputs and exact settings. Individual claims	Enhanced prediction of protein functional identity through the integration of sequence and structural features Original source ↗ Materials and methods/Prediction with ESM-2 (paragraph 2); Materials and methods/Prediction with DeepFRI (paragraph 1) Version: PMC11609699.1 Retrieved: 2026-09-16T10:33:35.728Z	source checked automated source review · 2026-09-16 Audit details Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: `attributes.profile.diagram.caption` Source artifact SHA-256: `db33e0542005ffae00cd644dfe697185b94c8823d5aee2768620a6db0c48e56f` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Diagram steps Pairs of protein amino-acid sequences ESM2 Pairwise cosine similarity used as a same-function score Individual claims	Enhanced prediction of protein functional identity through the integration of sequence and structural features Original source ↗ Materials and methods/Prediction with ESM-2 (paragraph 2); Materials and methods/Prediction with DeepFRI (paragraph 1) Version: PMC11609699.1 Retrieved: 2026-09-16T10:33:35.728Z	source checked automated source review · 2026-09-16 Audit details Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: `attributes.profile.diagram.steps` Source artifact SHA-256: `db33e0542005ffae00cd644dfe697185b94c8823d5aee2768620a6db0c48e56f` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Diagram title Evaluated procedure (conceptual) Individual claims	Enhanced prediction of protein functional identity through the integration of sequence and structural features Original source ↗ Materials and methods/Prediction with ESM-2 (paragraph 2); Materials and methods/Prediction with DeepFRI (paragraph 1) Version: PMC11609699.1 Retrieved: 2026-09-16T10:33:35.728Z	source checked automated source review · 2026-09-16 Audit details Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: `attributes.profile.diagram.title` Source artifact SHA-256: `db33e0542005ffae00cd644dfe697185b94c8823d5aee2768620a6db0c48e56f` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Model type Protein sequence transformer; this record is the paper-specific evaluated configuration. Individual claims	facebookresearch/esm README.md Original source ↗ README.md model description Version: 2b369911bb5b4b0dda914521b9475cad1656b2ac Retrieved: 2026-09-16T20:00:00.816433+00:00	source checked automated source review · 2026-09-16 Audit details Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: `attributes.profile.facts.0.value` Source artifact SHA-256: `8b273c21a322fc9473d1b68d0dd40c8166ab2f89e4a190aa26ca87251b97cba9` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Architecture / procedure esm2_t33_650M_UR50D emits 1,280-dimensional residue vectors. Mean pooling gives a protein vector; cosine similarity estimates whether two proteins share an enzymatic function. Individual claims	Enhanced prediction of protein functional identity through the integration of sequence and structural features Original source ↗ Materials and methods/Prediction with ESM-2 (paragraph 2); Materials and methods/Prediction with DeepFRI (paragraph 1) Version: PMC11609699.1 Retrieved: 2026-09-16T10:33:35.728Z	source checked automated source review · 2026-09-16 Audit details Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: `attributes.profile.facts.1.value` Source artifact SHA-256: `db33e0542005ffae00cd644dfe697185b94c8823d5aee2768620a6db0c48e56f` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Weights licence The inspected model-access documentation does not explicitly identify terms for this exact evaluated checkpoint or fitted head; repository code terms are shown separately. Individual claims	facebookresearch/esm README.md Original source ↗ README.md; checkpoint/access documentation and licence scope Version: 2b369911bb5b4b0dda914521b9475cad1656b2ac Retrieved: 2026-09-16T20:00:00.816433+00:00	unreported automated source review · 2026-09-16 Audit details Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: `attributes.profile.facts.10.value` Source artifact SHA-256: `8b273c21a322fc9473d1b68d0dd40c8166ab2f89e4a190aa26ca87251b97cba9` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Biological inputs Pairs of protein amino-acid sequences Individual claims	Enhanced prediction of protein functional identity through the integration of sequence and structural features Original source ↗ Materials and methods/Prediction with ESM-2 (paragraph 1); Introduction (paragraph 4) Version: PMC11609699.1 Retrieved: 2026-09-16T10:33:35.728Z	source checked automated source review · 2026-09-16 Audit details Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: `attributes.profile.facts.2.value` Source artifact SHA-256: `db33e0542005ffae00cd644dfe697185b94c8823d5aee2768620a6db0c48e56f` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Outputs Pairwise cosine similarity used as a same-function score Individual claims	Enhanced prediction of protein functional identity through the integration of sequence and structural features Original source ↗ Materials and methods/Prediction with ESM-2 (paragraph 2); Materials and methods/Prediction with DeepFRI (paragraph 1) Version: PMC11609699.1 Retrieved: 2026-09-16T10:33:35.728Z	source checked automated source review · 2026-09-16 Audit details Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: `attributes.profile.facts.3.value` Source artifact SHA-256: `db33e0542005ffae00cd644dfe697185b94c8823d5aee2768620a6db0c48e56f` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Parameters 650 million backbone parameters; no FUJISAN LightGBM features are part of this comparator. Individual claims	Enhanced prediction of protein functional identity through the integration of sequence and structural features Original source ↗ Results and discussion/Analysis of feature importance (paragraph 1); Materials and methods/Model training and hyperparameter optimization (paragraph 1) Version: PMC11609699.1 Retrieved: 2026-09-16T10:33:35.728Z	source checked automated source review · 2026-09-16 Audit details Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: `attributes.profile.facts.4.value` Source artifact SHA-256: `db33e0542005ffae00cd644dfe697185b94c8823d5aee2768620a6db0c48e56f` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Known versions / configuration esm2_t33_650M_UR50D Individual claims	Enhanced prediction of protein functional identity through the integration of sequence and structural features Original source ↗ Model identification in the comparison table and corresponding Methods; immutable checkpoint revision is not supplied by the table label. Version: PMC11609699.1 Retrieved: 2026-09-16T10:33:35.728Z	unreported automated source review · 2026-09-16 Audit details Primary full text and the available official implementation/model documentation were inspected. Explanatory claims are source-backed; unresolved exact-configuration metadata is labelled explicitly. This is automated review, not a human review or independent benchmark reproduction. Field: `attributes.profile.facts.5.value` Source artifact SHA-256: `db33e0542005ffae00cd644dfe697185b94c8823d5aee2768620a6db0c48e56f` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact

All evidence (CSV)All evidence (JSONL)How to read the evidence

Sources and history

View linked audit checks and correction history

Release 2026-09-29-06401fd5b220 · Record review: needs review

3 source records and release history

Enhanced prediction of protein functional identity through the integration of sequence and structural features · Original source · PMC11609699.1
facebookresearch/esm README.md · Original source · 2b369911bb5b4b0dda914521b9475cad1656b2ac
facebookresearch/esm LICENSE · Original source · 2b369911bb5b4b0dda914521b9475cad1656b2ac

Download this release

Technical metadata and extraction receipts

Stable ID: reported-model-ccd1160ad4ec27

areas: proteins-complexes
entity level: method
version: Not reported
reported name: ESM2
historical missing metadata: version: not_reported_in_legacy_extract; checkpoint revision: not_reported_in_legacy_extract; training data: not_reported_in_legacy_extract; licence: not_reported_in_legacy_extract
metadata review scope: historical_missing_metadata preserves the original discovery state. Current descriptive evidence and missingness are recorded in profile.facts; numerical-result review is separate.
legacy kinds: model
entity classification: review date: 2026-09-17; rationale: This source-scoped entry preserves the method/configuration actually named in an evaluation. It is neither a global family identity nor proof of an immutable checkpoint; the linked evaluation retains adaptation, fitting and scoring details.; source ids: fujisan-2024; evidence-reported-base-esm-readme-md; source locator: Materials and methods/Prediction with ESM-2 (paragraph 2); Materials and methods/Prediction with DeepFRI (paragraph 1) | README.md model description | Materials and methods/Prediction with ESM-2 (paragraph 2); Results and discussion/Limitations and perspectives (paragraph 1); ambiguities: Configuration means the source-labelled evaluated identity. It does not establish missing checkpoint hashes, default settings or equivalence to same-named records in other papers.

Related records

uses model: ESM-2
model: ESM2: Enzyme functional identity prediction
subject: ESM2: uses model ESM-2

Suggest a correction