Pipeline

MegSite + ESM3

MegSite predicts protein residues that bind DNA or RNA using multimodal ESM3 information.

Sources

MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Bidirectional cross-attention fusion/The architecture of MegSite (paragraph 1); Bidirectional cross-attention fusion/Implementation details (paragraph 2)

1 evaluation · 1 result

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

MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Results/Performance of different network architectures (paragraph 7); Materials and methods/Multimodal protein language model (paragraph 2)

Overview

Model type

Study-specific predictive method; this record is the paper-specific evaluated configuration.

Sources

Biological inputs

Protein sequence, structure and function representations

Sources

MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Results/Comparison with alternative protein language model embeddings (paragraph 3); Protein graph representation/Node feature (paragraph 1)

Outputs

Per-residue DNA- or RNA-binding predictions

Sources

MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Results/Comparison with state-of-the-art methods (paragraph 3); Protein graph representation (paragraph 1)

Access

Official study implementation and usage documentation: https://github.com/pengsl-lab/MegSite/blob/4d1f5441f15bd20eb3e7c9e5be2d4ca5a857f46b/README.md. This pinned documentation revision is not automatically the evaluated weight revision.

Sources

pengsl-lab/MegSite 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 · 1 result. 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
Pipeline: MegSite + ESM3	Task: DNA-binding residue prediction Dataset: DNA-129_Test	0.948 AUC fraction · unknown Uncertainty: Not reported Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source MegSite + ESM3: DNA-binding residue prediction ESM3 multimodal embedding ablation in MegSite Aggregation: Not reported MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Table 2, DNA-129_Test / ESM3 row, AUC column

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

Use this model

How it works, versions and access

Underlying model: ESM3. Results on this page belong to this pipeline and its evaluated settings.

How it works

How the evaluated method works

Sequence, structural and functional representations from ESM3 inform a task-specific nucleic-acid-binding-site predictor.

Sources

What was evaluated

The linked evaluation record identifies MegSite + ESM3: DNA-binding residue prediction. Its dataset, split, adaptation and evidence origin remain attached to the reported results.

Sources

MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · The named evaluation’s methods and comparison table; exact preserved evaluation IDs: evaluation-b2-megsite-2025

Strengths, limitations and unresolved questions

Strengths and limitations

Strengths and considerations

Includes independent tests and analyses of proteins with low structural similarity.
Sources
MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Results/Performance of different network architectures (paragraph 6); Abstract (paragraph 1)

Limitations and conditions

These are downstream binding-site predictions with multimodal inputs, not raw ESM3 outputs or evidence that predicted binding is experimentally realised.
Sources
MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Bidirectional cross-attention fusion/The architecture of MegSite (paragraph 1); Results/Comparison with alternative protein language model embeddings (paragraph 3)

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-86393c76dd8fa9

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	Study-specific predictive method; this record is the paper-specific evaluated configuration. Sources MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Results/Performance of different network architectures (paragraph 7); Materials and methods/Multimodal protein language model (paragraph 2)
Architecture / procedure	Sequence, structural and functional representations from ESM3 inform a task-specific nucleic-acid-binding-site predictor. Sources MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Results/Performance of different network architectures (paragraph 7); Materials and methods/Multimodal protein language model (paragraph 2)
Biological inputs	Protein sequence, structure and function representations Sources MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Results/Comparison with alternative protein language model embeddings (paragraph 3); Protein graph representation/Node feature (paragraph 1)
Outputs	Per-residue DNA- or RNA-binding predictions Sources MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Results/Comparison with state-of-the-art methods (paragraph 3); Protein graph representation (paragraph 1)
Parameters	The feature extractor is ESM3-Small: 1.4 billion parameters and 48 transformer blocks. The equivariant graph predictor is an additional fitted component. Sources MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Materials and methods/Multimodal protein language model (paragraph 2); Protein graph representation/Node feature (paragraph 1)
Known versions / configuration	MegSite + ESM3 is the comparison-table label; that label does not specify an immutable weight revision. · Not reported in inspected sources Sources MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Model identification in the comparison table and corresponding Methods; immutable checkpoint revision is not supplied by the table label.
Training data / fitting	DNA-573_Train and RNA-495_Train with independent DNA/RNA test sets; the paper filters training/test sequence similarity at 30%. Sources MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Materials and methods/Benchmark datasets (paragraph 1); Results/Comparison with state-of-the-art methods (paragraph 1)
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) MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model; pengsl-lab/MegSite README.md · Materials and methods/Benchmark datasets; Materials and methods/Multimodal protein language model; Bidirectional cross-attention fusion/Implementation details; Bidirectional cross-attention fusion/The architecture of MegSite; Results/Performance of different network architectures; Results/Comparison with state-of-the-art methods; inspected for explicit maximum input length (dataset lengths and family-wide limits are not substituted); README.md at pinned repository revision
Access	Official study implementation and usage documentation: https://github.com/pengsl-lab/MegSite/blob/4d1f5441f15bd20eb3e7c9e5be2d4ca5a857f46b/README.md. This pinned documentation revision is not automatically the evaluated weight revision. Sources pengsl-lab/MegSite README.md · README.md; installation, model download and usage instructions
Code licence	No explicit code licence was established from the paper’s availability statement and inspected repository-root documentation. · Not reported in inspected sources Sources pengsl-lab/MegSite README.md · README.md and repository-root licence-file search
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 pengsl-lab/MegSite 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.

21 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	MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model Original source ↗ Results/Performance of different network architectures (paragraph 7); Materials and methods/Multimodal protein language model (paragraph 2) Version: journal full text in PMC Retrieved: 2026-09-16T10:38:57.558213+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.diagram.caption` Source artifact SHA-256: `10d13122331813243d83b84fe6f9294eac7e7c03cde082ebed276191ac41089c` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Diagram steps Protein sequence, structure and function representations MegSite + ESM3 Per-residue DNA- or RNA-binding predictions Individual claims	MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model Original source ↗ Results/Performance of different network architectures (paragraph 7); Materials and methods/Multimodal protein language model (paragraph 2) Version: journal full text in PMC Retrieved: 2026-09-16T10:38:57.558213+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.diagram.steps` Source artifact SHA-256: `10d13122331813243d83b84fe6f9294eac7e7c03cde082ebed276191ac41089c` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Diagram title Evaluated procedure (conceptual) Individual claims	MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model Original source ↗ Results/Performance of different network architectures (paragraph 7); Materials and methods/Multimodal protein language model (paragraph 2) Version: journal full text in PMC Retrieved: 2026-09-16T10:38:57.558213+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.diagram.title` Source artifact SHA-256: `10d13122331813243d83b84fe6f9294eac7e7c03cde082ebed276191ac41089c` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Model type Study-specific predictive method; this record is the paper-specific evaluated configuration. Individual claims	MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model Original source ↗ Results/Performance of different network architectures (paragraph 7); Materials and methods/Multimodal protein language model (paragraph 2) Version: journal full text in PMC Retrieved: 2026-09-16T10:38:57.558213+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: `10d13122331813243d83b84fe6f9294eac7e7c03cde082ebed276191ac41089c` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Architecture / procedure Sequence, structural and functional representations from ESM3 inform a task-specific nucleic-acid-binding-site predictor. Individual claims	MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model Original source ↗ Results/Performance of different network architectures (paragraph 7); Materials and methods/Multimodal protein language model (paragraph 2) Version: journal full text in PMC Retrieved: 2026-09-16T10:38:57.558213+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.1.value` Source artifact SHA-256: `10d13122331813243d83b84fe6f9294eac7e7c03cde082ebed276191ac41089c` 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	pengsl-lab/MegSite README.md Original source ↗ README.md; checkpoint/access documentation and licence scope Version: 4d1f5441f15bd20eb3e7c9e5be2d4ca5a857f46b Retrieved: 2026-09-16T19:54:18.210404+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: `226fee71219189aded7aa165f7de40c85d2f96f4106089a8cd047793b524d18c` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Biological inputs Protein sequence, structure and function representations Individual claims	MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model Original source ↗ Results/Comparison with alternative protein language model embeddings (paragraph 3); Protein graph representation/Node feature (paragraph 1) Version: journal full text in PMC Retrieved: 2026-09-16T10:38:57.558213+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.2.value` Source artifact SHA-256: `10d13122331813243d83b84fe6f9294eac7e7c03cde082ebed276191ac41089c` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Outputs Per-residue DNA- or RNA-binding predictions Individual claims	MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model Original source ↗ Results/Comparison with state-of-the-art methods (paragraph 3); Protein graph representation (paragraph 1) Version: journal full text in PMC Retrieved: 2026-09-16T10:38:57.558213+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.3.value` Source artifact SHA-256: `10d13122331813243d83b84fe6f9294eac7e7c03cde082ebed276191ac41089c` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Parameters The feature extractor is ESM3-Small: 1.4 billion parameters and 48 transformer blocks. The equivariant graph predictor is an additional fitted component. Individual claims	MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model Original source ↗ Materials and methods/Multimodal protein language model (paragraph 2); Protein graph representation/Node feature (paragraph 1) Version: journal full text in PMC Retrieved: 2026-09-16T10:38:57.558213+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.4.value` Source artifact SHA-256: `10d13122331813243d83b84fe6f9294eac7e7c03cde082ebed276191ac41089c` Hash scope: Hash scope not separately documented; inspect source record Inspected artifact
Known versions / configuration MegSite + ESM3 is the comparison-table label; that label does not specify an immutable weight revision. Individual claims	MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model Original source ↗ Model identification in the comparison table and corresponding Methods; immutable checkpoint revision is not supplied by the table label. Version: journal full text in PMC Retrieved: 2026-09-16T10:38:57.558213+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.5.value` Source artifact SHA-256: `10d13122331813243d83b84fe6f9294eac7e7c03cde082ebed276191ac41089c` 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

2 source records and release history

MegSite: an accurate nucleic acid-binding residue prediction method based on multimodal protein language model · Original source · journal full text in PMC
pengsl-lab/MegSite README.md · Original source · 4d1f5441f15bd20eb3e7c9e5be2d4ca5a857f46b

Download this release

Technical metadata and extraction receipts

Stable ID: reported-model-86393c76dd8fa9

areas: proteins-complexes
entity level: method
version: not stated in table
reported name: MegSite + ESM3
historical missing metadata: 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 record identifies a composed analysis workflow with separately identifiable upstream models, representations or tools and a downstream prediction/scoring procedure. Results belong to that complete composition rather than to an upstream model alone.; source ids: megsite-2025; source locator: Results/Performance of different network architectures (paragraph 7); Materials and methods/Multimodal protein language model (paragraph 2) | Bidirectional cross-attention fusion/The architecture of MegSite (paragraph 1); Bidirectional cross-attention fusion/Implementation details (paragraph 2); ambiguities: This is the paper-specific pipeline identity; unspecified component checkpoints or implementation versions are not inferred from its name.

Related records

uses model: ESM3
model: MegSite + ESM3: DNA-binding residue prediction
subject: MegSite + ESM3: uses model ESM3

Suggest a correction