Protocol

Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1)

Zero-shot classification of pathogenic/likely pathogenic versus benign/likely benign variants in the ClinVar coding SNVs, poorly conserved stratum, AUROC and AUPRC.

18 evaluations · 36 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.

18 recorded evaluations, 36 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

18 evaluations · 36 results. Different protocols are not a single leaderboard.

Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) · Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)

Sorted by AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) (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	AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved)	AUROC (zero-shot; ClinVar coding SNVs, poorly conserved)
AlphaMissense (Evo 2 paper variant effect evaluations)Evaluation	0.883 (best in column)	0.957 (best in column)
ESM-1b (Evo 2 paper variant effect evaluations)Evaluation	0.853	0.937
CADD (Evo 2 paper variant effect evaluations)Evaluation	0.759	0.912
GPN-MSA (Evo 2 paper variant effect evaluations)Evaluation	0.708	0.882
Evo 2 7B base (Evo 2 paper variant effect evaluations)Evaluation	0.635	0.874
Evo 2 7B (Evo 2 paper variant effect evaluations)Evaluation	0.612	0.863
Evo 2 40B (Evo 2 paper variant effect evaluations)Evaluation	0.602	0.86
Evo 2 40B base (Evo 2 paper variant effect evaluations)Evaluation	0.593	0.854
ESM-2 650M (Evo 2 paper variant effect evaluations)Evaluation	0.503	0.696
DeCodon 200M (Evo 2 paper variant effect evaluations)Evaluation	0.379	0.641
EnCodon 620M (Evo 2 paper variant effect evaluations)Evaluation	0.346	0.597
PhyloP 241 (Evo 2 paper variant effect evaluations)Evaluation	0.311	0.576
PhyloP 447 (Evo 2 paper variant effect evaluations)Evaluation	0.305	0.578
PhyloP 470 (Evo 2 paper variant effect evaluations)Evaluation	0.265	0.537
CodonBERT 87M (Evo 2 paper variant effect evaluations)Evaluation	0.231	0.576
PhyloP 100 (Evo 2 paper variant effect evaluations)Evaluation	0.224	0.531
NT (Evo 2 paper variant effect evaluations)Evaluation	0.193	0.526
Evo 1 (Evo 2 paper variant effect evaluations)Evaluation	0.192	0.531

All 36 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: AlphaMissense (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.883 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source AlphaMissense on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J425, row 425: Model 'AlphaMissense', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: AlphaMissense (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.957 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source AlphaMissense on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I425, row 425: Model 'AlphaMissense', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: CADD (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.759 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source CADD on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J427, row 427: Model 'CADD', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: CADD (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.912 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source CADD on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I427, row 427: Model 'CADD', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: CodonBERT 87M (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.231 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source CodonBERT 87M on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J438, row 438: Model 'CodonBERT 87M', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: CodonBERT 87M (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.576 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source CodonBERT 87M on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I438, row 438: Model 'CodonBERT 87M', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: DeCodon 200M (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.379 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source DeCodon 200M on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J434, row 434: Model 'DeCodon 200M', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: DeCodon 200M (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.641 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source DeCodon 200M on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I434, row 434: Model 'DeCodon 200M', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: EnCodon 620M (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.346 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source EnCodon 620M on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J435, row 435: Model 'EnCodon 620M', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: EnCodon 620M (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.597 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source EnCodon 620M on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I435, row 435: Model 'EnCodon 620M', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: ESM-1b (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.853 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-1b on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J426, row 426: Model 'ESM-1b', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: ESM-1b (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.937 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-1b on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I426, row 426: Model 'ESM-1b', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: ESM-2 650M (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.503 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-2 650M on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J433, row 433: Model 'ESM-2 650M', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: ESM-2 650M (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.696 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source ESM-2 650M on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I433, row 433: Model 'ESM-2 650M', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: Evo 1 (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.192 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source Evo 1 on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J440, row 440: Model 'Evo 1', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: Evo 1 (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.531 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source Evo 1 on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I440, row 440: Model 'Evo 1', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: Evo 2 40B base (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.593 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source Evo 2 40B base on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J432, row 432: Model 'Evo 2 40B base', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: Evo 2 40B base (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.854 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source Evo 2 40B base on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I432, row 432: Model 'Evo 2 40B base', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: Evo 2 40B (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.602 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source Evo 2 40B on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J431, row 431: Model 'Evo 2 40B', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: Evo 2 40B (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.86 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source Evo 2 40B on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I431, row 431: Model 'Evo 2 40B', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: Evo 2 7B base (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.635 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source Evo 2 7B base on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J429, row 429: Model 'Evo 2 7B base', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: Evo 2 7B base (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.874 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source Evo 2 7B base on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I429, row 429: Model 'Evo 2 7B base', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: Evo 2 7B (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.612 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source Evo 2 7B on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J430, row 430: Model 'Evo 2 7B', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'
Configuration: Evo 2 7B (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.863 AUROC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source Evo 2 7B on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell I430, row 430: Model 'Evo 2 7B', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUROC'
Configuration: GPN-MSA (Evo 2 paper variant effect evaluations)	Protocol: Zero-shot ClinVar coding SNVs, poorly conserved classification (Evo 2 paper, Supplementary Data 1) Dataset: ClinVar October 2024 release, filtered (Brixi et al. 2026)	0.708 AUPRC (zero-shot; ClinVar coding SNVs, poorly conserved) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source GPN-MSA on ClinVar coding SNVs, poorly conserved (Evo 2 paper) gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved Aggregation: Not reported Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026); Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Supplementary Data 1 (sheet 'Sheet1') cell J428, row 428: Model 'GPN-MSA', Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved', column 'AUPRC'

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.

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.

Author-reported evaluations: 18

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

Null control

Proposed control: requires review

Protocol-valid seeded random ranking; no fitting on assay labels

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

Established label-free reference with explicitly permitted sequence/MSA inputs

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.

Strengths, limitations and unresolved questions

Evidence

Source checking verifies the cited claim or transcription. It does not establish independent reproduction.

Evidence table

Inspect claims, sources and review details

Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.

One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.

0 evidence rows matching the loaded filters

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

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

How to read the evidence

Sources and history

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

3 source records and release history

evo2: Journal full-text XML · Original source · Retrieved page snapshot; no immutable publisher revision supplied
Genome modelling and design across all domains of life with Evo 2: Supplementary Information (Nature 2026) · Original source · Nature 652 (2026), published 2026-03-04; Supplementary Information file 41586_2026_10176_MOESM1_ESM.pdf (73 pages)
Genome modelling and design across all domains of life with Evo 2: Supplementary Data 1, human variant effect prediction results (Nature 2026) · Original source · Nature 652 (2026), published 2026-03-04; Supplementary Data 1 file 41586_2026_10176_MOESM3_ESM.xlsx

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: gfm-guides-20261014-evo2-brixi2026-protocol-clinvar-coding-snv-poorly-conserved

areas: dna-genomes
contexts: research
protocol: Score each variant zero-shot (for Evo 2, the log-likelihood of the variant sequence minus that of the reference over an 8,192-bp genomic window around the variant) and compute AUROC and AUPRC for pathogenic/likely pathogenic versus benign/likely benign variants in the stratum Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved'.
version: As published, Nature 652 (2026), Supplementary Data 1
source locator: Supplementary Data 1 rows with Data 'ClinVar', Genomic_Feature 'Coding', Variant_Type 'SNV', Sub_category 'Poorly conserved'; SI A.3.12 and A.3.13 (page 13); Extended Data Fig. 4
limitations: Positives: 146 pathogenic/likely pathogenic; negatives: 636 benign/likely benign; total 782 variants (Supplementary Data 1).; Label 'Poorly conserved'; SI A.3.13 names its bins 'weakly conserved' (0 <= phyloP < 0.5) and 'accelerated evolution' (phyloP < 0); the mapping is not stated.; No split: every model scores the same variants zero-shot (SI A.3.12: 'For each benchmark, we score the same variants with all models').; Supervised comparators (AlphaMissense, CADD, SpliceAI, Pangolin) may have been trained on overlapping variants; training overlap is not reported.; No uncertainty is reported.

Related records

part of: Evo 2 paper zero-shot human variant effect evaluations (Brixi et al. 2026)
uses data: ClinVar October 2024 release, filtered (Brixi et al. 2026)
assessment: AlphaMissense on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: CADD on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: CodonBERT 87M on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: DeCodon 200M on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: EnCodon 620M on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: ESM-1b on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: ESM-2 650M on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: Evo 1 on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: Evo 2 40B base on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: Evo 2 40B on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: Evo 2 7B base on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: Evo 2 7B on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: GPN-MSA on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: NT on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: PhyloP 100 on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: PhyloP 241 on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: PhyloP 447 on ClinVar coding SNVs, poorly conserved (Evo 2 paper)
assessment: PhyloP 470 on ClinVar coding SNVs, poorly conserved (Evo 2 paper)

Suggest a correction