Configuration

NGNN with GCN combined with SEAL with randomly initialised node features (baseline) (PPI embeddings study)

Evidence concern: excluded from comparisons

The text's summary does not match the tables: it lists ProtT5 among structure embeddings while the tables print ProstT5; it reports a GraphSAGE with DistMult result (+38% Hits@100 on HuRI) although DistMult is not a column; and the stated gains (for example ESM-2 with GraphSAGE +6% Hits@100 on OGBl-ppa) cannot be reproduced from the printed table values (GS random 0.0104, ESM-2 0.2421). Values are kept as printed; comparisons should rely on the tables only after a reviewer has judged this. Random-feature baselines on OGBl-ppa (GCN 0.0588, GS 0.0104 Hits@100) look far below the public OGB leaderboard values the authors say they reproduced (Supplementary Table S2, not retrieved; reviewer did not verify the leaderboard values).The role of protein embeddings for protein-protein interaction prediction with graph neural networks · Results and discussion, first paragraph; Tables 5 and 6

NGNN with GCN combined with SEAL with randomly initialised node features (baseline): NGNN with GCN combined with SEAL; node features: randomly initialised node features (baseline).

4 evaluations · 4 results

Overview

An explanatory profile has not yet been reviewed for this record.

Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.

Evaluations and results

4 evaluations · 4 results. Different protocols are not a single leaderboard.

Sorted by Accuracy (HuRI PPI link prediction) (higher is better). The best value in each column is highlighted. Decimals are rounded for display; each value links to the printed value and its source.

Tested configuration	Protocol and dataset	Accuracy (HuRI PPI link prediction)	Accuracy (OGBl-ppa PPI link prediction)	Hits at 100 (HuRI PPI link prediction)	Hits at 100 (OGBl-ppa PPI link prediction)
NGNN with GCN combined with SEAL with randomly initialised node features (baseline) (PPI embeddings study)Evaluation	HuRI PPI link prediction, Accuracy (Table 6), HuRI PPI graph (embedding study)	0.753	Not reported	Not reported	Not reported
NGNN with GCN combined with SEAL with randomly initialised node features (baseline) (PPI embeddings study)Evaluation	HuRI PPI link prediction, Hits@100 (Table 5), HuRI PPI graph (embedding study)	Not reported	Not reported	0.646	Not reported
NGNN with GCN combined with SEAL with randomly initialised node features (baseline) (PPI embeddings study)Evaluation	OGBl-ppa PPI link prediction, Accuracy (Table 6), OGBl-ppa PPI graph (embedding study)	Not reported	0.896	Not reported	Not reported
NGNN with GCN combined with SEAL with randomly initialised node features (baseline) (PPI embeddings study)Evaluation	OGBl-ppa PPI link prediction, Hits@100 (Table 5), OGBl-ppa PPI graph (embedding study)	Not reported	Not reported	Not reported	0.512

All 4 result rows with coverage, uncertainty and sources

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: NGNN with GCN combined with SEAL with randomly initialised node features (baseline) (PPI embeddings study)	Protocol: HuRI PPI link prediction, Accuracy (Table 6) Dataset: HuRI PPI graph (embedding study)	0.753 Accuracy (HuRI PPI link prediction) fraction · higher Uncertainty: Not reported by the source: Only a table-wide maximum SD is printed (Table 5/6 footnote) Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons The text's summary does not match the tables: it lists ProtT5 among structure embeddings while the tables print ProstT5; it reports a GraphSAGE with DistMult result (+38% Hits@100 on HuRI) although DistMult is not a column; and the stated gains (for example ESM-2 with GraphSAGE +6% Hits@100 on OGBl-ppa) cannot be reproduced from the printed table values (GS random 0.0104, ESM-2 0.2421). Values are kept as printed; comparisons should rely on the tables only after a reviewer has judged this. Random-feature baselines on OGBl-ppa (GCN 0.0588, GS 0.0104 Hits@100) look far below the public OGB leaderboard values the authors say they reproduced (Supplementary Table S2, not retrieved; reviewer did not verify the leaderboard values).The role of protein embeddings for protein-protein interaction prediction with graph neural networks · Results and discussion, first paragraph; Tables 5 and 6 Methods, coverage and source NGNN with GCN combined with SEAL / Random on HuRI (Accuracy) lit-interactions-20261011-ppiemb-protocol-huri-accuracy Aggregation: Not reported The role of protein embeddings for protein-protein interaction prediction with graph neural networks · Table 6, block 'HuRI', row 'N-GCN+S', column 'Random'
Configuration: NGNN with GCN combined with SEAL with randomly initialised node features (baseline) (PPI embeddings study)	Protocol: HuRI PPI link prediction, Hits@100 (Table 5) Dataset: HuRI PPI graph (embedding study)	0.646 Hits at 100 (HuRI PPI link prediction) fraction · higher Uncertainty: Not reported by the source: Only a table-wide maximum SD is printed (Table 5/6 footnote) Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons The text's summary does not match the tables: it lists ProtT5 among structure embeddings while the tables print ProstT5; it reports a GraphSAGE with DistMult result (+38% Hits@100 on HuRI) although DistMult is not a column; and the stated gains (for example ESM-2 with GraphSAGE +6% Hits@100 on OGBl-ppa) cannot be reproduced from the printed table values (GS random 0.0104, ESM-2 0.2421). Values are kept as printed; comparisons should rely on the tables only after a reviewer has judged this. Random-feature baselines on OGBl-ppa (GCN 0.0588, GS 0.0104 Hits@100) look far below the public OGB leaderboard values the authors say they reproduced (Supplementary Table S2, not retrieved; reviewer did not verify the leaderboard values).The role of protein embeddings for protein-protein interaction prediction with graph neural networks · Results and discussion, first paragraph; Tables 5 and 6 Methods, coverage and source NGNN with GCN combined with SEAL / Random on HuRI (Hits@100) lit-interactions-20261011-ppiemb-protocol-huri-hits-100 Aggregation: Not reported The role of protein embeddings for protein-protein interaction prediction with graph neural networks · Table 5, block 'HuRI', row 'N-GCN+S', column 'Random'
Configuration: NGNN with GCN combined with SEAL with randomly initialised node features (baseline) (PPI embeddings study)	Protocol: OGBl-ppa PPI link prediction, Accuracy (Table 6) Dataset: OGBl-ppa PPI graph (embedding study)	0.896 Accuracy (OGBl-ppa PPI link prediction) fraction · higher Uncertainty: Not reported by the source: Only a table-wide maximum SD is printed (Table 5/6 footnote) Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons The text's summary does not match the tables: it lists ProtT5 among structure embeddings while the tables print ProstT5; it reports a GraphSAGE with DistMult result (+38% Hits@100 on HuRI) although DistMult is not a column; and the stated gains (for example ESM-2 with GraphSAGE +6% Hits@100 on OGBl-ppa) cannot be reproduced from the printed table values (GS random 0.0104, ESM-2 0.2421). Values are kept as printed; comparisons should rely on the tables only after a reviewer has judged this. Random-feature baselines on OGBl-ppa (GCN 0.0588, GS 0.0104 Hits@100) look far below the public OGB leaderboard values the authors say they reproduced (Supplementary Table S2, not retrieved; reviewer did not verify the leaderboard values).The role of protein embeddings for protein-protein interaction prediction with graph neural networks · Results and discussion, first paragraph; Tables 5 and 6 Methods, coverage and source NGNN with GCN combined with SEAL / Random on OGBl-ppa (Accuracy) lit-interactions-20261011-ppiemb-protocol-ogbl-ppa-accuracy Aggregation: Not reported The role of protein embeddings for protein-protein interaction prediction with graph neural networks · Table 6, block 'OGBl-ppa', row 'N-GCN+S', column 'Random'
Configuration: NGNN with GCN combined with SEAL with randomly initialised node features (baseline) (PPI embeddings study)	Protocol: OGBl-ppa PPI link prediction, Hits@100 (Table 5) Dataset: OGBl-ppa PPI graph (embedding study)	0.512 Hits at 100 (OGBl-ppa PPI link prediction) fraction · higher Uncertainty: Not reported by the source: Only a table-wide maximum SD is printed (Table 5/6 footnote) Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons The text's summary does not match the tables: it lists ProtT5 among structure embeddings while the tables print ProstT5; it reports a GraphSAGE with DistMult result (+38% Hits@100 on HuRI) although DistMult is not a column; and the stated gains (for example ESM-2 with GraphSAGE +6% Hits@100 on OGBl-ppa) cannot be reproduced from the printed table values (GS random 0.0104, ESM-2 0.2421). Values are kept as printed; comparisons should rely on the tables only after a reviewer has judged this. Random-feature baselines on OGBl-ppa (GCN 0.0588, GS 0.0104 Hits@100) look far below the public OGB leaderboard values the authors say they reproduced (Supplementary Table S2, not retrieved; reviewer did not verify the leaderboard values).The role of protein embeddings for protein-protein interaction prediction with graph neural networks · Results and discussion, first paragraph; Tables 5 and 6 Methods, coverage and source NGNN with GCN combined with SEAL / Random on OGBl-ppa (Hits@100) lit-interactions-20261011-ppiemb-protocol-ogbl-ppa-hits-100 Aggregation: Not reported The role of protein embeddings for protein-protein interaction prediction with graph neural networks · Table 5, block 'OGBl-ppa', row 'N-GCN+S', column 'Random'

Source checking is not independent reproduction. Release 2026-10-22-1b7cf027e4f6.

Use this configuration

How it works, versions and access

Strengths, limitations and unresolved questions

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.

0 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-10-22-1b7cf027e4f6
Property and statement	Original source and location	Review and provenance

No evidence rows match these filters. Choose another scope or clear the search.

How to read the evidence

Sources and history

Release 2026-10-22-1b7cf027e4f6 · Record review: source checked

1 source record and release history

The role of protein embeddings for protein-protein interaction prediction with graph neural networks · Original source · Briefings in Bioinformatics 2026, bbag313; PMC13280947 full-text XML

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: lit-interactions-20261011-ppiemb-config-n-gcn-plus-s-random

areas: molecular-interactions; proteins-complexes
contexts: research
method types: supervised_machine_learning
reported name: N-GCN+S / Random
foundation model eligible: false
parameters: NGNN with GCN combined with SEAL; node features: randomly initialised node features (baseline). Leaderboard hyperparameters reproduced for OGBl-ppa (Experimental design).
source locator: Tables 5 and 6, row 'N-GCN+S', column 'Random'
access reuse: Not stated in the source
missing metadata: version: reason: unreported; note: Embedding and GNN parameters are in Supplementary Tables S3, S5-S7, not retrieved
provenance note: NGNN with GCN combined with SEAL using randomly initialised node features (baseline), as evaluated for PPI prediction.

Related records

configuration of: NGNN with GCN combined with SEAL
system: NGNN with GCN combined with SEAL / Random on HuRI (Accuracy)
system: NGNN with GCN combined with SEAL / Random on HuRI (Hits@100)
system: NGNN with GCN combined with SEAL / Random on OGBl-ppa (Accuracy)
system: NGNN with GCN combined with SEAL / Random on OGBl-ppa (Hits@100)

Suggest a correction