ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2); Supplementary Table 2 sheet 'Table_of_results_stop_gains'

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.

10 recorded evaluations, 10 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

10 evaluations · 10 results. Different protocols are not a single leaderboard.

Protocol: ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2)

Sorted by AUROC (ROC-AUC, pathogenic versus benign ClinVar all stop-gains; zero-shot) (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	AUROC (ROC-AUC, pathogenic versus benign ClinVar all stop-gains; zero-shot)	AUROC (ROC-AUC, pathogenic versus benign ClinVar stop-gains expected to undergo NMD (50 bp rule); zero-shot)	AUROC (ROC-AUC, pathogenic versus benign ClinVar stop-gains not expected to undergo NMD (50 bp rule); zero-shot)
Number of residues lost (stop-gain baseline)Evaluation	ClinVar stop-gain variants, pathogenic vs benign (Brandes et al. 2023)	0.695 (best in column)	Not reported	Not reported
ESM1b minimum-LLR stop-gain score (Brandes et al. 2023)Evaluation	ClinVar stop-gain variants, pathogenic vs benign (Brandes et al. 2023)	0.683	Not reported	Not reported
50 bp rule for nonsense-mediated decayEvaluation	ClinVar stop-gain variants, pathogenic vs benign (Brandes et al. 2023)	0.662	Not reported	Not reported
Percentage of residues lost (stop-gain baseline)Evaluation	ClinVar stop-gain variants, pathogenic vs benign (Brandes et al. 2023)	0.579	Not reported	Not reported
ESM1b minimum-LLR stop-gain score (Brandes et al. 2023)Evaluation	ClinVar stop-gains expected to undergo NMD (Brandes et al. 2023)	Not reported	0.566	Not reported
Number of residues lost (stop-gain baseline)Evaluation	ClinVar stop-gains expected to undergo NMD (Brandes et al. 2023)	Not reported	0.59 (best in column)	Not reported
Percentage of residues lost (stop-gain baseline)Evaluation	ClinVar stop-gains expected to undergo NMD (Brandes et al. 2023)	Not reported	0.452	Not reported
ESM1b minimum-LLR stop-gain score (Brandes et al. 2023)Evaluation	ClinVar stop-gains not expected to undergo NMD (Brandes et al. 2023)	Not reported	Not reported	0.734 (best in column)
Number of residues lost (stop-gain baseline)Evaluation	ClinVar stop-gains not expected to undergo NMD (Brandes et al. 2023)	Not reported	Not reported	0.65
Percentage of residues lost (stop-gain baseline)Evaluation	ClinVar stop-gains not expected to undergo NMD (Brandes et al. 2023)	Not reported	Not reported	0.6

All 10 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: ESM1b minimum-LLR stop-gain score (Brandes et al. 2023)	Protocol: ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2) Dataset: ClinVar stop-gain variants, pathogenic vs benign (Brandes et al. 2023)	0.683 AUROC (ROC-AUC, pathogenic versus benign ClinVar all stop-gains; zero-shot) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM1b-minLLR on all stop-gains (Brandes et al. 2023) gfm-guides-20261014-esm1b-protocol-clinvar-stop-gains Aggregation: Not reported Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · Supplementary Table 2 sheet 'Table_of_results_stop_gains' cell D2, row 'ESM1b-minLLR', column 'AUC_ALL'
Method: 50 bp rule for nonsense-mediated decay	Protocol: ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2) Dataset: ClinVar stop-gain variants, pathogenic vs benign (Brandes et al. 2023)	0.662 AUROC (ROC-AUC, pathogenic versus benign ClinVar all stop-gains; zero-shot) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source 50bp rule on all stop-gains (Brandes et al. 2023) gfm-guides-20261014-esm1b-protocol-clinvar-stop-gains Aggregation: Not reported Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · Supplementary Table 2 sheet 'Table_of_results_stop_gains' cell D5, row '50bp rule', column 'AUC_ALL'
Configuration: ESM1b minimum-LLR stop-gain score (Brandes et al. 2023)	Protocol: ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2) Dataset subset: ClinVar stop-gains expected to undergo NMD (Brandes et al. 2023)	0.566 AUROC (ROC-AUC, pathogenic versus benign ClinVar stop-gains expected to undergo NMD (50 bp rule); zero-shot) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM1b-minLLR on stop-gains expected to undergo NMD (50 bp rule) (Brandes et al. 2023) gfm-guides-20261014-esm1b-protocol-clinvar-stop-gains Aggregation: Not reported Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · Supplementary Table 2 sheet 'Table_of_results_stop_gains' cell C2, row 'ESM1b-minLLR', column 'AUC_NMD'
Method: Number of residues lost (stop-gain baseline)	Protocol: ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2) Dataset subset: ClinVar stop-gains expected to undergo NMD (Brandes et al. 2023)	0.59 AUROC (ROC-AUC, pathogenic versus benign ClinVar stop-gains expected to undergo NMD (50 bp rule); zero-shot) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source num. AA lost on stop-gains expected to undergo NMD (50 bp rule) (Brandes et al. 2023) gfm-guides-20261014-esm1b-protocol-clinvar-stop-gains Aggregation: Not reported Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · Supplementary Table 2 sheet 'Table_of_results_stop_gains' cell C3, row 'num. AA lost', column 'AUC_NMD'
Method: Percentage of residues lost (stop-gain baseline)	Protocol: ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2) Dataset subset: ClinVar stop-gains expected to undergo NMD (Brandes et al. 2023)	0.452 AUROC (ROC-AUC, pathogenic versus benign ClinVar stop-gains expected to undergo NMD (50 bp rule); zero-shot) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source pct. AA lost on stop-gains expected to undergo NMD (50 bp rule) (Brandes et al. 2023) gfm-guides-20261014-esm1b-protocol-clinvar-stop-gains Aggregation: Not reported Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · Supplementary Table 2 sheet 'Table_of_results_stop_gains' cell C4, row 'pct. AA lost', column 'AUC_NMD'
Configuration: ESM1b minimum-LLR stop-gain score (Brandes et al. 2023)	Protocol: ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2) Dataset subset: ClinVar stop-gains not expected to undergo NMD (Brandes et al. 2023)	0.734 AUROC (ROC-AUC, pathogenic versus benign ClinVar stop-gains not expected to undergo NMD (50 bp rule); zero-shot) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM1b-minLLR on stop-gains not expected to undergo NMD (50 bp rule) (Brandes et al. 2023) gfm-guides-20261014-esm1b-protocol-clinvar-stop-gains Aggregation: Not reported Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · Supplementary Table 2 sheet 'Table_of_results_stop_gains' cell B2, row 'ESM1b-minLLR', column 'AUC_no_NMD'
Method: Number of residues lost (stop-gain baseline)	Protocol: ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2) Dataset subset: ClinVar stop-gains not expected to undergo NMD (Brandes et al. 2023)	0.65 AUROC (ROC-AUC, pathogenic versus benign ClinVar stop-gains not expected to undergo NMD (50 bp rule); zero-shot) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source num. AA lost on stop-gains not expected to undergo NMD (50 bp rule) (Brandes et al. 2023) gfm-guides-20261014-esm1b-protocol-clinvar-stop-gains Aggregation: Not reported Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · Supplementary Table 2 sheet 'Table_of_results_stop_gains' cell B3, row 'num. AA lost', column 'AUC_no_NMD'
Method: Percentage of residues lost (stop-gain baseline)	Protocol: ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2) Dataset subset: ClinVar stop-gains not expected to undergo NMD (Brandes et al. 2023)	0.6 AUROC (ROC-AUC, pathogenic versus benign ClinVar stop-gains not expected to undergo NMD (50 bp rule); zero-shot) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source pct. AA lost on stop-gains not expected to undergo NMD (50 bp rule) (Brandes et al. 2023) gfm-guides-20261014-esm1b-protocol-clinvar-stop-gains Aggregation: Not reported Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · Supplementary Table 2 sheet 'Table_of_results_stop_gains' cell B4, row 'pct. AA lost', column 'AUC_no_NMD'
Method: Number of residues lost (stop-gain baseline)	Protocol: ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2) Dataset: ClinVar stop-gain variants, pathogenic vs benign (Brandes et al. 2023)	0.695 AUROC (ROC-AUC, pathogenic versus benign ClinVar all stop-gains; zero-shot) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source num. AA lost on all stop-gains (Brandes et al. 2023) gfm-guides-20261014-esm1b-protocol-clinvar-stop-gains Aggregation: Not reported Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · Supplementary Table 2 sheet 'Table_of_results_stop_gains' cell D3, row 'num. AA lost', column 'AUC_ALL'
Method: Percentage of residues lost (stop-gain baseline)	Protocol: ClinVar stop-gain pathogenicity by NMD status, ROC-AUC (Brandes et al. 2023 Fig. 6d, Supplementary Table 2) Dataset: ClinVar stop-gain variants, pathogenic vs benign (Brandes et al. 2023)	0.579 AUROC (ROC-AUC, pathogenic versus benign ClinVar all stop-gains; zero-shot) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source pct. AA lost on all stop-gains (Brandes et al. 2023) gfm-guides-20261014-esm1b-protocol-clinvar-stop-gains Aggregation: Not reported Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · Supplementary Table 2 sheet 'Table_of_results_stop_gains' cell D4, row 'pct. AA lost', column 'AUC_ALL'

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

Methods and evaluation design

Procedure, tasks and evaluated configurations

Recorded evaluations

Each evaluation records what was tested and under which conditions.

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.

Author-reported evaluations: 10

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 2026-10-22-1b7cf027e4f6. 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 2026-10-22-1b7cf027e4f6
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

2 source records and release history

Genome-wide prediction of disease variant effects with a deep protein language model (Brandes et al., Nature Genetics 2023) · Original source · Version of record, Nature Genetics 55, 1512-1522 (published 2023-08-10)
Brandes et al. 2023 Supplementary Table 2: full benchmark evaluations of ESM1b and all other methods · Original source · Version of record, Nature Genetics 55, 1512-1522 (published 2023-08-10)

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: gfm-guides-20261014-esm1b-protocol-clinvar-stop-gains

areas: dna-genomes; protein-function
contexts: clinical_research
protocol: Scores used directly (no training on these labels). The ESM1b stop-gain score is the lowest missense LLR in the region lost after the new stop codon; baselines are the number and percentage of residues lost and the binary 50 bp NMD rule.
version: Supplementary Table 2 sheet 'Table_of_results_stop_gains'
metric: auroc
metric direction: higher
unit: unitless
limitations: Benchmark built by the ESM1b workflow's authors (author-reported for ESM1b and the baselines).; Classes are highly imbalanced (345 benign of 36,379).; Fig. 6d plots the mean of 20 bootstrap iterations, each sampling 140 pathogenic and 140 benign variants; whether the workbook values are those means or ROC-AUCs on the whole sets is not stated.; The 50 bp rule gives binary labels, so only its all-variant ROC-AUC is printed.; Review 2026-10-14: each workbook stop-gain value times twice the product of the class counts printed in the Methods (3,672 x 147 without NMD; 32,441 x 198 with NMD; 36,113 x 345 overall) is an integer, as a ROC-AUC over the whole set is; with the legend counts (32,362 and 36,034 pathogenic) it is not, and a mean of 20 bootstrap ROC-AUCs on 140 + 140 variants would not be. The workbook values therefore appear to be whole-set ROC-AUCs on the Methods counts, not the plotted bootstrap means; the source does not say so.
source locator: Results 'ESM1b can predict the effects of multiresidue variants', paragraph beginning 'Stop-gain variant effects can be predicted'; Methods 'Generalized effect scores for indels and stop-gain variants', 'Baseline scores of indel and stop-gain variant effects' and 'AUC metrics for pathogenicity classification' (stop-gain bootstrap paragraph); Supplementary Table 2 'Contents'

Related records