result · source checked
0.89 AUROC
scRegNet (Geneformer backbone) · AUROC · hESC cell-type-specific GRN
- Tested model
- scRegNet (Geneformer backbone)
- Task or benchmark
- Gene-regulatory link prediction
- Dataset
- hESC cell-type-specific GRN
- Procedure
- TFs plus 500 variable genes; mean from 50 independent evaluations.
- Evaluation
- scRegNet (Geneformer backbone): Gene-regulatory link prediction
- Evidence
- Author-reported evaluation · source checkedPrediction of Gene Regulatory Connections with Joint Single-Cell Foundation Models and Graph-Based Learning · Table 2, scRegNet (w/ Geneformer) row, hESC AUROC entry
A source-checked result verifies the numerical transcription, not every model or protocol detail. Evaluation metadata: needs review. Source checked does not mean independently reproduced.
Evaluation results
Release 2026-09-16-d74d282221a9 · 1 evaluation · 1 metric row. Different protocols are not a single leaderboard.
| Metric and finding | Coverage and uncertainty | Evidence |
|---|---|---|
| scRegNet (Geneformer backbone): Gene-regulatory link prediction Model: scRegNet (Geneformer backbone) · Benchmark: Gene-regulatory link prediction · Dataset: hESC cell-type-specific GRN TFs plus 500 variable genes; mean from 50 independent evaluations. Author-reported evaluation · Evaluation metadata: needs review | ||
| 0.89 AUROC Unit: unitless · Direction: unknown | Uncertainty: ± 0.00 as printed Scored: Not reported · Eligible: Not reported | source checkedPrediction of Gene Regulatory Connections with Joint Single-Cell Foundation Models and Graph-Based Learning · Table 2, scRegNet (w/ Geneformer) row, hESC AUROC entry Source checking is not independent reproduction. |
Sources and history
Release 2026-09-16-d74d282221a9 · Record review: source checked
- Prediction of Gene Regulatory Connections with Joint Single-Cell Foundation Models and Graph-Based Learning · Original source · PMC11838224.2
Technical metadata and extraction receipts
Stable ID: lit-031
- areas
- cells-tissues
- tasks
- Gene-regulatory link prediction
- printed value
- 0.89
- numeric value
- 0.89
- metric
- AUROC
- metric direction
- unknown
- unit
- unitless
- uncertainty
- ± 0.00 as printed
- source locator
- Table 2, scRegNet (w/ Geneformer) row, hESC AUROC entry
- review
- method: independent_ai_table_review; reviewer: Codex omics research agent; independent source-table review, not human review; reviewed at: 2026-09-16T10:41:16.541287+00:00; notes: Read break elements: cells contain AUROC on first line then AUPRC. Selected hESC (first cell type), first line. Caption specifies 500 most-variable genes and 50 independent evaluations. This verifies the central score at its source location, not every metadata field or an experimental reproduction.; evidence: {"table_xml_id": "T2", "row_cells": ["scRegNet (w/ Geneformer)", "AUROCAUPRC", "0.89±0.000.62±0.00", "0.90±0.000.84±0.00", "0.81±0.000.17±0.00", "0.93±0.000.86±0.00", "0.92±0.000.94±0.00", "0.93±0.000.94±0.00", "0.88±0.000.88±0.00"], "selected_cell_zero_based": 2, "selected_cell_xml": "<td align=\"center\" valign=\"top\" rowspan=\"1\" colspan=\"1\"><bold>0.89</bold>±0.00<break /><bold>0.62</bold>±0.00</td>", "caption": "Link prediction performance on seven scRNA-seq datasets with 500 most-variable genes. Each dataset includes a cell-type-specific ground-truth network. The values reported are averages from 50 independent evaluations per cell type. scRegNet utilizing the three backbone models—scBERT, Geneformer, and scFoundation—consistently outperforms the baselines."}; artifact sha256: 65b3272d47bb9c4ee1e7a965169bef63add9dbeb31508d4076e5145b761af4ec; retrieval url: https://www.ebi.ac.uk/europepmc/webservices/rest/PMC11838224/fullTextXML
- legacy id
- lit-031
- legacy row
- id: lit-031; paper id: scregnet-2025; domain id: cells-tissues; task: Gene-regulatory link prediction; model: scRegNet (Geneformer backbone); model version: Not reported; dataset: hESC cell-type-specific GRN; dataset version: Not reported; split: Not reported; metric: AUROC; value: 0.89; unit: unitless; uncertainty: ± 0.00 as printed; protocol: TFs plus 500 variable genes; mean from 50 independent evaluations.; source locator: Table 2, scRegNet (w/ Geneformer) row, hESC AUROC entry; source url: https://pmc.ncbi.nlm.nih.gov/articles/PMC11838224/; evaluation origin: author_reported; reviewed utc: 2026-09-15T23:25:00Z
- missing metadata
- model version: not_reported_in_legacy_extract; dataset version: not_reported_in_legacy_extract; split: not_reported_in_legacy_extract