Protocol

Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference

Evidence concern: excluded from comparisons

Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4

Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption

Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio

Rank TF-target candidate edges by absolute weight; early precision ratio scores the top K edges (K = number of reference edges) and AUPRC ratio the full ranked list, each over a random predictor, against the STRING functional interactions reference.

147 evaluations · 294 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.

147 recorded evaluations, 294 metric rows. A comparison chart has not yet been validated for these results. The table retains the individual findings and their sources.

View coverage and remaining gaps across all benchmarks

Results

Results are available, but no reviewed comparison panel is linked in this release.

All evaluations

147 evaluations · 294 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: DeepSEM (Ye et al. 2026 baseline)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	2.1 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source DeepSEM GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'ML', column 'DeepSEM'
Configuration: DeepSEM (Ye et al. 2026 baseline)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	4.14 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source DeepSEM GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'ML', column 'DeepSEM'
Configuration: Geneformer attn GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.77 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source Geneformer attn GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'attn', column 'Geneformer'
Configuration: Geneformer attn GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	3.27 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source Geneformer attn GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'attn', column 'Geneformer'
Configuration: Geneformer coshid GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.1 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source Geneformer coshid GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'coshid', column 'Geneformer'
Configuration: Geneformer coshid GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.3 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source Geneformer coshid GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'coshid', column 'Geneformer'
Configuration: Geneformer costok GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.11 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source Geneformer costok GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'costok', column 'Geneformer'
Configuration: Geneformer costok GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.34 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source Geneformer costok GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'costok', column 'Geneformer'
Configuration: GENIE3 (Ye et al. 2026 baseline)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.62 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source GENIE3 GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'ML', column 'GENIE3'
Configuration: GENIE3 (Ye et al. 2026 baseline)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	2.91 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source GENIE3 GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'ML', column 'GENIE3'
Configuration: GRNBoost (Ye et al. 2026 baseline)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.17 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source GRNboost GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'ML', column 'GRNboost'
Configuration: GRNBoost (Ye et al. 2026 baseline)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	2.81 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source GRNboost GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'ML', column 'GRNboost'
Configuration: LangCell attn GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.86 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source LangCell attn GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'attn', column 'Langcell'
Configuration: LangCell attn GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	3.3 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source LangCell attn GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'attn', column 'Langcell'
Configuration: LangCell coshid GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.4 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source LangCell coshid GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'coshid', column 'Langcell'
Configuration: LangCell coshid GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	2.39 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source LangCell coshid GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'coshid', column 'Langcell'
Configuration: LangCell costok GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.11 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source LangCell costok GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'costok', column 'Langcell'
Configuration: LangCell costok GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.38 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source LangCell costok GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'costok', column 'Langcell'
Configuration: scCello attn GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.26 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source scCello attn GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'attn', column 'scCello'
Configuration: scCello attn GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.34 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source scCello attn GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'attn', column 'scCello'
Configuration: scCello coshid GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.49 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source scCello coshid GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'coshid', column 'scello'
Configuration: scCello coshid GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	2.39 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source scCello coshid GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'coshid', column 'scCello'
Configuration: scCello costok GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.18 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source scCello costok GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'costok', column 'scello'
Configuration: scCello costok GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.94 Early precision ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source scCello costok GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 4, block 'STRING', row 'hESC', column group 'costok', column 'scCello'
Configuration: scFoundation attn GRN (Ye et al. 2026)	Protocol: Ye et al. 2026 GRN reconstruction, TFs+500, STRING reference Dataset subset: BEELINE hESC, TFs+500 (Ye et al. 2026)	1.05 AUPRC ratio (TFs+500 gene set; STRING reference) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Evidence concern: excluded from comparisons Table 1 labels its middle reference column 'ChIP-seq' and the caption names cell-type-specific ChIP-seq, but its values equal the 'Nonspecific ChIP-seq' mean and std rows of Tables 4 and 6 (for example scGPT costok EPR 4.211±1.703), and Appendix B says cell-type-specific ChIP-seq is excluded from the main comparison. Table 1 is therefore not extracted; Tables 4-7 carry the reference labels.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 1 caption and column header; Tables 4 and 6, 'Nonspecific ChIP-seq' mean rows; Appendix B text before Table 4 Table 6 is titled 'AUPR' and Table 7 'AUPRC', but both print ratios over a random predictor (values above 1), matching the 'AUPRC ratio' of Table 1; both are recorded as auprc-ratio.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6 and Table 7 captions; Table 1 caption Table 7 (TFs+1000 AUPRC) repeats Table 6 (TFs+500 AUPR) exactly in the GENIE3 column of the STRING, Cell-type-specific ChIP-seq and Nonspecific ChIP-seq blocks: all 18 dataset values and the three mean and std rows are identical (for example STRING hESC 1.6241, mean 3.301, std 2.244). Only the OmniPath block differs. Every other column, and GENIE3 in the EPR Tables 4 and 5, differs between the two gene sets, and the Table 6 values match Table 1. These 21 Table 7 GENIE3 values are likely a copy of the TFs+500 values and should not be compared.Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 7, column group 'ML', column 'GENIE3', blocks 'STRING', 'Cell-type-specific' and 'Nonspecific'; Table 6, same cells; Table 1, GENIE3 AUPRC ratio Methods, coverage and source scFoundation attn GRN on hESC, TFs+500, STRING lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string Aggregation: Not reported Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Table 6, block 'STRING', row 'hESC', column group 'attn', column 'scFoundation'

Source checking is not independent reproduction. Release 2026-10-23-5347695500ed.

Methods and evaluation design

Procedure, tasks and evaluated configurations

Recorded evaluations

Each evaluation records what was tested and under which conditions.

Explore all linked results

Baseline coverage

Reference methods help show what a model adds beyond simple controls. We track a null control and a conventional method for each protocol.

0 of 2 active baseline roles have published Rewire measurements in this release. Measurements on a selected protocol do not establish coverage of an entire suite.

No execution recipe linked to this protocol. Recipe availability does not establish a completed evaluation.

External evaluations: 147

Literature evidence is not a Rewire measurement. Executed but unpublished runs and private review status are not included.

Null control

Proposed control: requires review

Select a task-valid null control after reviewing inputs and metric

Protocol-specific applicability, permitted inputs, access, split, evaluator and execution requirements need review before implementation or execution.

This is a suggested selection rule, not a validated method or a measured score.

Conventional reference

Proposed control: requires review

Select an upstream conventional reference after reviewing the full protocol

Protocol-specific applicability, permitted inputs, access, split, evaluator and execution requirements need review before implementation or execution.

This is a suggested selection rule, not a validated method or a measured score.

Protocol coverage CSV (gzip) · Model evaluation matrix (gzip) · Source table (gzip) · Release and checksums (gzip)

Coverage is derived from release 2026-10-23-5347695500ed. Source citations describe the original records; they do not validate an unreviewed baseline proposal. No results have been generated by this audit.

Run instructions

No runnable recipe has been reviewed for this protocol. Dataset access, model requirements, licences and compute requirements must be checked against its sources before execution.

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 2026-10-23-5347695500ed
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 2026-10-23-5347695500ed · Record review: source checked

1 source record and release history

Benchmarking Static Gene Regulatory Network Reconstruction and Dynamic Transition Probing in Single-Cell Foundation Models · Original source · bioRxiv preprint version 1, posted 2026-05-20

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: lit-cells-networks-20261011-ye2026grn-protocol-tf-plus-500-string

areas: biological-networks; single-cell
contexts: research
protocol: TFs+500 gene set; candidate space restricted to reference TFs as sources and reference-set genes as targets; STRING reference network; early precision and precision-recall over ranked edges (Appendix B.2).
version: bioRxiv v1
metric: early-precision-ratio
metric direction: higher
unit: unitless
limitations: One run per dataset; dataset-level spread only.; Ratios over a random predictor depend on reference network density, so values are not comparable across references or gene sets.
source locator: Tables 4-7; Section 2.2; Appendix B.2

Related records

part of: Static GRN reconstruction from single-cell foundation models (Ye et al. 2026)
assessment: DeepSEM GRN on hESC, TFs+500, STRING
assessment: Geneformer attn GRN on hESC, TFs+500, STRING
assessment: Geneformer coshid GRN on hESC, TFs+500, STRING
assessment: Geneformer costok GRN on hESC, TFs+500, STRING
assessment: GENIE3 GRN on hESC, TFs+500, STRING
assessment: GRNboost GRN on hESC, TFs+500, STRING
assessment: LangCell attn GRN on hESC, TFs+500, STRING
assessment: LangCell coshid GRN on hESC, TFs+500, STRING
assessment: LangCell costok GRN on hESC, TFs+500, STRING
assessment: scCello attn GRN on hESC, TFs+500, STRING
assessment: scCello coshid GRN on hESC, TFs+500, STRING
assessment: scCello costok GRN on hESC, TFs+500, STRING
assessment: scFoundation attn GRN on hESC, TFs+500, STRING
assessment: scFoundation coshid GRN on hESC, TFs+500, STRING
assessment: scFoundation costok GRN on hESC, TFs+500, STRING
assessment: scGPT attn GRN on hESC, TFs+500, STRING
assessment: scGPT coshid GRN on hESC, TFs+500, STRING
assessment: scGPT costok GRN on hESC, TFs+500, STRING
assessment: scPRINT attn GRN on hESC, TFs+500, STRING
assessment: scPRINT coshid GRN on hESC, TFs+500, STRING
assessment: scPRINT costok GRN on hESC, TFs+500, STRING
assessment: DeepSEM GRN on hHep, TFs+500, STRING
assessment: Geneformer attn GRN on hHep, TFs+500, STRING
assessment: Geneformer coshid GRN on hHep, TFs+500, STRING
assessment: Geneformer costok GRN on hHep, TFs+500, STRING
assessment: GENIE3 GRN on hHep, TFs+500, STRING
assessment: GRNboost GRN on hHep, TFs+500, STRING
assessment: LangCell attn GRN on hHep, TFs+500, STRING
assessment: LangCell coshid GRN on hHep, TFs+500, STRING
assessment: LangCell costok GRN on hHep, TFs+500, STRING
assessment: scCello attn GRN on hHep, TFs+500, STRING
assessment: scCello coshid GRN on hHep, TFs+500, STRING
assessment: scCello costok GRN on hHep, TFs+500, STRING
assessment: scFoundation attn GRN on hHep, TFs+500, STRING
assessment: scFoundation coshid GRN on hHep, TFs+500, STRING
assessment: scFoundation costok GRN on hHep, TFs+500, STRING
assessment: scGPT attn GRN on hHep, TFs+500, STRING
assessment: scGPT coshid GRN on hHep, TFs+500, STRING
assessment: scGPT costok GRN on hHep, TFs+500, STRING
assessment: scPRINT attn GRN on hHep, TFs+500, STRING
assessment: scPRINT coshid GRN on hHep, TFs+500, STRING
assessment: scPRINT costok GRN on hHep, TFs+500, STRING
assessment: DeepSEM GRN on mDC, TFs+500, STRING
assessment: Geneformer attn GRN on mDC, TFs+500, STRING
assessment: Geneformer coshid GRN on mDC, TFs+500, STRING
assessment: Geneformer costok GRN on mDC, TFs+500, STRING
assessment: GENIE3 GRN on mDC, TFs+500, STRING
assessment: GRNboost GRN on mDC, TFs+500, STRING
assessment: LangCell attn GRN on mDC, TFs+500, STRING
assessment: LangCell coshid GRN on mDC, TFs+500, STRING
assessment: LangCell costok GRN on mDC, TFs+500, STRING
assessment: scCello attn GRN on mDC, TFs+500, STRING
assessment: scCello coshid GRN on mDC, TFs+500, STRING
assessment: scCello costok GRN on mDC, TFs+500, STRING
assessment: scFoundation attn GRN on mDC, TFs+500, STRING
assessment: scFoundation coshid GRN on mDC, TFs+500, STRING
assessment: scFoundation costok GRN on mDC, TFs+500, STRING
assessment: scGPT attn GRN on mDC, TFs+500, STRING
assessment: scGPT coshid GRN on mDC, TFs+500, STRING
assessment: scGPT costok GRN on mDC, TFs+500, STRING
assessment: scPRINT attn GRN on mDC, TFs+500, STRING
assessment: scPRINT coshid GRN on mDC, TFs+500, STRING
assessment: scPRINT costok GRN on mDC, TFs+500, STRING
assessment: DeepSEM GRN on mHSC-E, TFs+500, STRING
assessment: Geneformer attn GRN on mHSC-E, TFs+500, STRING
assessment: Geneformer coshid GRN on mHSC-E, TFs+500, STRING
assessment: Geneformer costok GRN on mHSC-E, TFs+500, STRING
assessment: GENIE3 GRN on mHSC-E, TFs+500, STRING
assessment: GRNboost GRN on mHSC-E, TFs+500, STRING
assessment: LangCell attn GRN on mHSC-E, TFs+500, STRING
assessment: LangCell coshid GRN on mHSC-E, TFs+500, STRING
assessment: LangCell costok GRN on mHSC-E, TFs+500, STRING
assessment: scCello attn GRN on mHSC-E, TFs+500, STRING
assessment: scCello coshid GRN on mHSC-E, TFs+500, STRING
assessment: scCello costok GRN on mHSC-E, TFs+500, STRING
assessment: scFoundation attn GRN on mHSC-E, TFs+500, STRING
assessment: scFoundation coshid GRN on mHSC-E, TFs+500, STRING
assessment: scFoundation costok GRN on mHSC-E, TFs+500, STRING
assessment: scGPT attn GRN on mHSC-E, TFs+500, STRING
assessment: scGPT coshid GRN on mHSC-E, TFs+500, STRING
assessment: scGPT costok GRN on mHSC-E, TFs+500, STRING
assessment: scPRINT attn GRN on mHSC-E, TFs+500, STRING
assessment: scPRINT coshid GRN on mHSC-E, TFs+500, STRING
assessment: scPRINT costok GRN on mHSC-E, TFs+500, STRING
assessment: DeepSEM GRN on mHSC-GM, TFs+500, STRING
assessment: Geneformer attn GRN on mHSC-GM, TFs+500, STRING
assessment: Geneformer coshid GRN on mHSC-GM, TFs+500, STRING
assessment: Geneformer costok GRN on mHSC-GM, TFs+500, STRING
assessment: GENIE3 GRN on mHSC-GM, TFs+500, STRING
assessment: GRNboost GRN on mHSC-GM, TFs+500, STRING
assessment: LangCell attn GRN on mHSC-GM, TFs+500, STRING
assessment: LangCell coshid GRN on mHSC-GM, TFs+500, STRING
assessment: LangCell costok GRN on mHSC-GM, TFs+500, STRING
assessment: scCello attn GRN on mHSC-GM, TFs+500, STRING
assessment: scCello coshid GRN on mHSC-GM, TFs+500, STRING
assessment: scCello costok GRN on mHSC-GM, TFs+500, STRING
assessment: scFoundation attn GRN on mHSC-GM, TFs+500, STRING
assessment: scFoundation coshid GRN on mHSC-GM, TFs+500, STRING
assessment: scFoundation costok GRN on mHSC-GM, TFs+500, STRING
assessment: scGPT attn GRN on mHSC-GM, TFs+500, STRING
assessment: scGPT coshid GRN on mHSC-GM, TFs+500, STRING
assessment: scGPT costok GRN on mHSC-GM, TFs+500, STRING
assessment: scPRINT attn GRN on mHSC-GM, TFs+500, STRING
assessment: scPRINT coshid GRN on mHSC-GM, TFs+500, STRING
assessment: scPRINT costok GRN on mHSC-GM, TFs+500, STRING
assessment: DeepSEM GRN on mHSC-L, TFs+500, STRING
assessment: Geneformer attn GRN on mHSC-L, TFs+500, STRING
assessment: Geneformer coshid GRN on mHSC-L, TFs+500, STRING
assessment: Geneformer costok GRN on mHSC-L, TFs+500, STRING
assessment: GENIE3 GRN on mHSC-L, TFs+500, STRING
assessment: GRNboost GRN on mHSC-L, TFs+500, STRING
assessment: LangCell attn GRN on mHSC-L, TFs+500, STRING
assessment: LangCell coshid GRN on mHSC-L, TFs+500, STRING
assessment: LangCell costok GRN on mHSC-L, TFs+500, STRING
assessment: scCello attn GRN on mHSC-L, TFs+500, STRING
assessment: scCello coshid GRN on mHSC-L, TFs+500, STRING
assessment: scCello costok GRN on mHSC-L, TFs+500, STRING
assessment: scFoundation attn GRN on mHSC-L, TFs+500, STRING
assessment: scFoundation coshid GRN on mHSC-L, TFs+500, STRING
assessment: scFoundation costok GRN on mHSC-L, TFs+500, STRING
assessment: scGPT attn GRN on mHSC-L, TFs+500, STRING
assessment: scGPT coshid GRN on mHSC-L, TFs+500, STRING
assessment: scGPT costok GRN on mHSC-L, TFs+500, STRING
assessment: scPRINT attn GRN on mHSC-L, TFs+500, STRING
assessment: scPRINT coshid GRN on mHSC-L, TFs+500, STRING
assessment: scPRINT costok GRN on mHSC-L, TFs+500, STRING
assessment: DeepSEM GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: Geneformer attn GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: Geneformer coshid GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: Geneformer costok GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: GENIE3 GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: GRNboost GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: LangCell attn GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: LangCell coshid GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: LangCell costok GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scCello attn GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scCello coshid GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scCello costok GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scFoundation attn GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scFoundation coshid GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scFoundation costok GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scGPT attn GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scGPT coshid GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scGPT costok GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scPRINT attn GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scPRINT coshid GRN on six BEELINE datasets (mean), TFs+500, STRING
assessment: scPRINT costok GRN on six BEELINE datasets (mean), TFs+500, STRING

Suggest a correction