GSMFormer-PPI + ProstT5: description
Description of GSMFormer-PPI + ProstT5
Predicts protein-protein interactions from three modalities: MaSIF surface descriptors, graph convolutions over the 3D structure, and ProstT5 residue embeddings as graph node features, fused by a transformer encoder.
Replace a placeholder description with a sourced functional description (issue #88). Historical releases retain the original value.
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.
7 evidence rows matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Claimed field description Context-only references | Multimodal graph, surface, and language-based model for protein protein interaction prediction Reviewed profile (2026-09-16): summary at Results/Ablation experiments for input modalities (paragraph 1); Conclusions (paragraph 1); architecture fact at Methods/Model architecture/Integration of protein structural and sequence features (paragraph 2); Methods/Model architecture/Linear projectors (paragraph 1); description text also read at Abstract; Methods/Model architecture (ProstT5 per-residue embeddings as node features) Version: journal full text in PMC | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Previous value Identity as reported in this paper. Unspecified versions are not assumed equivalent to other papers. Context-only references | Multimodal graph, surface, and language-based model for protein protein interaction prediction Reviewed profile (2026-09-16): summary at Results/Ablation experiments for input modalities (paragraph 1); Conclusions (paragraph 1); architecture fact at Methods/Model architecture/Integration of protein structural and sequence features (paragraph 2); Methods/Model architecture/Linear projectors (paragraph 1); description text also read at Abstract; Methods/Model architecture (ProstT5 per-residue embeddings as node features) Version: journal full text in PMC | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Source location Reviewed profile (2026-09-16): summary at Results/Ablation experiments for input modalities (paragraph 1); Conclusions (paragraph 1); architecture fact at Methods/Model architecture/Integration of protein structural and sequence features (paragraph 2); Methods/Model architecture/Linear projectors (paragraph 1); description text also read at Abstract; Methods/Model architecture (ProstT5 per-residue embeddings as node features) Context-only references | Multimodal graph, surface, and language-based model for protein protein interaction prediction Reviewed profile (2026-09-16): summary at Results/Ablation experiments for input modalities (paragraph 1); Conclusions (paragraph 1); architecture fact at Methods/Model architecture/Integration of protein structural and sequence features (paragraph 2); Methods/Model architecture/Linear projectors (paragraph 1); description text also read at Abstract; Methods/Model architecture (ProstT5 per-residue embeddings as node features) Version: journal full text in PMC | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Claimed value Predicts protein-protein interactions from three modalities: MaSIF surface descriptors, graph convolutions over the 3D structure, and ProstT5 residue embeddings as graph node features, fused by a transformer encoder. Context-only references | Multimodal graph, surface, and language-based model for protein protein interaction prediction Reviewed profile (2026-09-16): summary at Results/Ablation experiments for input modalities (paragraph 1); Conclusions (paragraph 1); architecture fact at Methods/Model architecture/Integration of protein structural and sequence features (paragraph 2); Methods/Model architecture/Linear projectors (paragraph 1); description text also read at Abstract; Methods/Model architecture (ProstT5 per-residue embeddings as node features) Version: journal full text in PMC | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Description Replace a placeholder description with a sourced functional description (issue #88). Historical releases retain the original value. Context-only references | Multimodal graph, surface, and language-based model for protein protein interaction prediction Reviewed profile (2026-09-16): summary at Results/Ablation experiments for input modalities (paragraph 1); Conclusions (paragraph 1); architecture fact at Methods/Model architecture/Integration of protein structural and sequence features (paragraph 2); Methods/Model architecture/Linear projectors (paragraph 1); description text also read at Abstract; Methods/Model architecture (ProstT5 per-residue embeddings as node features) Version: journal full text in PMC | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Relationship: subject reported-model-73ae07fb5be204 Context-only references | Multimodal graph, surface, and language-based model for protein protein interaction prediction Reviewed profile (2026-09-16): summary at Results/Ablation experiments for input modalities (paragraph 1); Conclusions (paragraph 1); architecture fact at Methods/Model architecture/Integration of protein structural and sequence features (paragraph 2); Methods/Model architecture/Linear projectors (paragraph 1); description text also read at Abstract; Methods/Model architecture (ProstT5 per-residue embeddings as node features) Version: journal full text in PMC | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
| Name GSMFormer-PPI + ProstT5: description Context-only references | Multimodal graph, surface, and language-based model for protein protein interaction prediction Reviewed profile (2026-09-16): summary at Results/Ablation experiments for input modalities (paragraph 1); Conclusions (paragraph 1); architecture fact at Methods/Model architecture/Integration of protein structural and sequence features (paragraph 2); Methods/Model architecture/Linear projectors (paragraph 1); description text also read at Abstract; Methods/Model architecture (ProstT5 per-residue embeddings as node features) Version: journal full text in PMC | not individually reviewed No individual claim review recorded Audit detailsField: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record |
Sources and history
Release 2026-11-21-c8bea25e5abb · Record review: source checked
1 source record and release history
- Multimodal graph, surface, and language-based model for protein protein interaction prediction · Original source · journal full text in PMC
Technical metadata and extraction receipts
Stable ID: metadata-correction-eb8765f46b9425040561
- areas
- proteins-complexes
- field
- description
- previous value
- Identity as reported in this paper. Unspecified versions are not assumed equivalent to other papers.
- value
- Predicts protein-protein interactions from three modalities: MaSIF surface descriptors, graph convolutions over the 3D structure, and ProstT5 residue embeddings as graph node features, fused by a transformer encoder.
- source locator
- Reviewed profile (2026-09-16): summary at Results/Ablation experiments for input modalities (paragraph 1); Conclusions (paragraph 1); architecture fact at Methods/Model architecture/Integration of protein structural and sequence features (paragraph 2); Methods/Model architecture/Linear projectors (paragraph 1); description text also read at Abstract; Methods/Model architecture (ProstT5 per-residue embeddings as node features)
- review
- method: automated-source-review; ai-assisted-source-review; reviewer: claude; reviewer note: Extractor: Claude integrator agent for issue #88. Reviewer: a separate Claude review agent, independent of the extractor; no human review claimed.; date: 2026-10-11; method note: Compared the store diff with origin/main, read each new description against its cited source passage and checked the licence of each new source record.; note: Approved. Checked all 11 scTab and MPRabc method descriptions against PMC11298532 and PMC13232518, and Lazypipe-nt, scaLR, CATHe2, DiffDock-NMDN, PC-mer and Hi-Enhancer against their papers; all match. The other pipeline descriptions are consistent with their pinned papers.; receipt: data/omics/qa-descriptions-20261011-c/review.json
Related records
- subject: GSMFormer-PPI + ProstT5