rewirebio.iobenchmarks
Protocol

ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class

Scores variants by the log likelihood ratio of the alternate over the reference allele and reports one AUROC per printed variant class.

60 evaluations · 60 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.

60 recorded evaluations, 60 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

60 evaluations · 60 results. Different protocols are not a single leaderboard.

Protocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class

Sorted by AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class 3’ UTR) (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 configurationProtocol and datasetAUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class 3’ UTR)AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class 5’ UTR)AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Downstream of Transcript)AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Intronic)AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Missense)AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Noncoding)AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Splice Acceptor)AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Splice Donor)AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Start Codon)AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Stop Lost)AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Synonymous)AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Upstream of Transcript)
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar 3’ UTR variants, at least one review star0.96 (best in column)Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar 3’ UTR variants, at least one review star0.94Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar 3’ UTR variants, at least one review star0.66Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar 3’ UTR variants, at least one review star0.63Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar 3’ UTR variants, at least one review star0.48Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar Downstream of Transcript variants, at least one review starNot reportedNot reported0.72Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar Downstream of Transcript variants, at least one review starNot reportedNot reported0.99 (best in column)Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar Downstream of Transcript variants, at least one review starNot reportedNot reported0.55Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar Downstream of Transcript variants, at least one review starNot reportedNot reported0.58Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar Downstream of Transcript variants, at least one review starNot reportedNot reported0.97Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar Intronic variants, at least one review starNot reportedNot reportedNot reported0.64Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar Intronic variants, at least one review starNot reportedNot reportedNot reported0.94Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar Intronic variants, at least one review starNot reportedNot reportedNot reported0.48Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar Intronic variants, at least one review starNot reportedNot reportedNot reported0.61Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar Intronic variants, at least one review starNot reportedNot reportedNot reported0.95 (best in column)Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar Missense variants, at least one review starNot reportedNot reportedNot reportedNot reported0.55Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar Missense variants, at least one review starNot reportedNot reportedNot reportedNot reported0.91 (best in column)Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar Missense variants, at least one review starNot reportedNot reportedNot reportedNot reported0.54Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar Missense variants, at least one review starNot reportedNot reportedNot reportedNot reported0.64Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar Missense variants, at least one review starNot reportedNot reportedNot reportedNot reported0.84Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar Noncoding variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reported0.59Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar Noncoding variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reported0.95 (best in column)Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar Noncoding variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reported0.47Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar Noncoding variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reported0.54Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar Noncoding variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reported0.94Not reportedNot reportedNot reportedNot reportedNot reportedNot reported
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar Splice Acceptor variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.61Not reportedNot reportedNot reportedNot reportedNot reported
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar Splice Acceptor variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.81Not reportedNot reportedNot reportedNot reportedNot reported
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar Splice Acceptor variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.51Not reportedNot reportedNot reportedNot reportedNot reported
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar Splice Acceptor variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.62Not reportedNot reportedNot reportedNot reportedNot reported
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar Splice Acceptor variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.82 (best in column)Not reportedNot reportedNot reportedNot reportedNot reported
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar Splice Donor variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.68Not reportedNot reportedNot reportedNot reported
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar Splice Donor variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.83 (best in column)Not reportedNot reportedNot reportedNot reported
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar Splice Donor variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.52Not reportedNot reportedNot reportedNot reported
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar Splice Donor variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.67Not reportedNot reportedNot reportedNot reported
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar Splice Donor variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.83 (best in column)Not reportedNot reportedNot reportedNot reported
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar Start Codon variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.41Not reportedNot reportedNot reported
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar Start Codon variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.8 (best in column)Not reportedNot reportedNot reported
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar Start Codon variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.43Not reportedNot reportedNot reported
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar Start Codon variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.69Not reportedNot reportedNot reported
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar Start Codon variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.77Not reportedNot reportedNot reported
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar Stop Lost variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.43Not reportedNot reported
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar Stop Lost variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.86 (best in column)Not reportedNot reported
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar Stop Lost variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.48Not reportedNot reported
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar Stop Lost variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.55Not reportedNot reported
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar Stop Lost variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.81Not reportedNot reported
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar Synonymous variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.55Not reported
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar Synonymous variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.87 (best in column)Not reported
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar Synonymous variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.43Not reported
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar Synonymous variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.55Not reported
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar Synonymous variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.85Not reported
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar Upstream of Transcript variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.4
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar Upstream of Transcript variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.84 (best in column)
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar Upstream of Transcript variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.51
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar Upstream of Transcript variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.31
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar Upstream of Transcript variants, at least one review starNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported0.74
Caduceus-PS 131k, 256 model dimensions, 16 layersEvaluationClinVar 5’ UTR variants, at least one review starNot reported0.61Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
GPN-MSA, published precomputed hg38 likelihood ratiosEvaluationClinVar 5’ UTR variants, at least one review starNot reported0.95 (best in column)Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
HyenaDNA medium 160k, likelihood-ratio scoringEvaluationClinVar 5’ UTR variants, at least one review starNot reported0.48Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringEvaluationClinVar 5’ UTR variants, at least one review starNot reported0.55Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
PhyloGPN (481 bp receptive field, about 83M parameters)EvaluationClinVar 5’ UTR variants, at least one review starNot reported0.94Not reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reportedNot reported
All 60 result rows with coverage, uncertainty and sources
Filter evaluations

Applied filters: All linked evaluations

Exact evaluated configurations and original reported results
Tested configurationProtocol and datasetFindingEvidence and details
Configuration: Caduceus-PS 131k, 256 model dimensions, 16 layersProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Downstream of Transcript variants, at least one review star
0.72 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Downstream of Transcript)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

Caduceus on ClinVar Downstream of Transcript variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Downstream of Transcript', column 'Caduceus'
Configuration: GPN-MSA, published precomputed hg38 likelihood ratiosProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Downstream of Transcript variants, at least one review star
0.99 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Downstream of Transcript)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

GPN-MSA on ClinVar Downstream of Transcript variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Downstream of Transcript', column 'GPN-MSA'
Configuration: HyenaDNA medium 160k, likelihood-ratio scoringProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Downstream of Transcript variants, at least one review star
0.55 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Downstream of Transcript)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

HyenaDNA on ClinVar Downstream of Transcript variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Downstream of Transcript', column 'HyenaDNA'
Configuration: Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Downstream of Transcript variants, at least one review star
0.58 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Downstream of Transcript)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

Nucleotide Transformer on ClinVar Downstream of Transcript variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Downstream of Transcript', column 'Nucleotide Transformer'
Configuration: PhyloGPN (481 bp receptive field, about 83M parameters)Protocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Downstream of Transcript variants, at least one review star
0.97 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Downstream of Transcript)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

PhyloGPN on ClinVar Downstream of Transcript variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Downstream of Transcript', column 'PhyloGPN'
Configuration: Caduceus-PS 131k, 256 model dimensions, 16 layersProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Intronic variants, at least one review star
0.64 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Intronic)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

Caduceus on ClinVar Intronic variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Intronic', column 'Caduceus'
Configuration: GPN-MSA, published precomputed hg38 likelihood ratiosProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Intronic variants, at least one review star
0.94 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Intronic)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

GPN-MSA on ClinVar Intronic variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Intronic', column 'GPN-MSA'
Configuration: HyenaDNA medium 160k, likelihood-ratio scoringProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Intronic variants, at least one review star
0.48 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Intronic)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

HyenaDNA on ClinVar Intronic variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Intronic', column 'HyenaDNA'
Configuration: Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Intronic variants, at least one review star
0.61 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Intronic)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

Nucleotide Transformer on ClinVar Intronic variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Intronic', column 'Nucleotide Transformer'
Configuration: PhyloGPN (481 bp receptive field, about 83M parameters)Protocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Intronic variants, at least one review star
0.95 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Intronic)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

PhyloGPN on ClinVar Intronic variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Intronic', column 'PhyloGPN'
Configuration: Caduceus-PS 131k, 256 model dimensions, 16 layersProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Missense variants, at least one review star
0.55 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Missense)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

Caduceus on ClinVar Missense variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Missense', column 'Caduceus'
Configuration: GPN-MSA, published precomputed hg38 likelihood ratiosProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Missense variants, at least one review star
0.91 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Missense)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

GPN-MSA on ClinVar Missense variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Missense', column 'GPN-MSA'
Configuration: HyenaDNA medium 160k, likelihood-ratio scoringProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Missense variants, at least one review star
0.54 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Missense)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

HyenaDNA on ClinVar Missense variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Missense', column 'HyenaDNA'
Configuration: Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Missense variants, at least one review star
0.64 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Missense)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

Nucleotide Transformer on ClinVar Missense variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Missense', column 'Nucleotide Transformer'
Configuration: PhyloGPN (481 bp receptive field, about 83M parameters)Protocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Missense variants, at least one review star
0.84 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Missense)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

PhyloGPN on ClinVar Missense variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Missense', column 'PhyloGPN'
Configuration: Caduceus-PS 131k, 256 model dimensions, 16 layersProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Noncoding variants, at least one review star
0.59 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Noncoding)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

Caduceus on ClinVar Noncoding variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Noncoding', column 'Caduceus'
Configuration: GPN-MSA, published precomputed hg38 likelihood ratiosProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Noncoding variants, at least one review star
0.95 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Noncoding)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

GPN-MSA on ClinVar Noncoding variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Noncoding', column 'GPN-MSA'
Configuration: HyenaDNA medium 160k, likelihood-ratio scoringProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Noncoding variants, at least one review star
0.47 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Noncoding)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

HyenaDNA on ClinVar Noncoding variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Noncoding', column 'HyenaDNA'
Configuration: Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Noncoding variants, at least one review star
0.54 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Noncoding)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

Nucleotide Transformer on ClinVar Noncoding variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Noncoding', column 'Nucleotide Transformer'
Configuration: PhyloGPN (481 bp receptive field, about 83M parameters)Protocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Noncoding variants, at least one review star
0.94 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Noncoding)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

PhyloGPN on ClinVar Noncoding variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Noncoding', column 'PhyloGPN'
Configuration: Caduceus-PS 131k, 256 model dimensions, 16 layersProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Splice Acceptor variants, at least one review star
0.61 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Splice Acceptor)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

Caduceus on ClinVar Splice Acceptor variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Splice Acceptor', column 'Caduceus'
Configuration: GPN-MSA, published precomputed hg38 likelihood ratiosProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Splice Acceptor variants, at least one review star
0.81 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Splice Acceptor)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

GPN-MSA on ClinVar Splice Acceptor variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Splice Acceptor', column 'GPN-MSA'
Configuration: HyenaDNA medium 160k, likelihood-ratio scoringProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Splice Acceptor variants, at least one review star
0.51 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Splice Acceptor)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

HyenaDNA on ClinVar Splice Acceptor variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Splice Acceptor', column 'HyenaDNA'
Configuration: Nucleotide Transformer v2 500M multi-species, likelihood-ratio scoringProtocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Splice Acceptor variants, at least one review star
0.62 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Splice Acceptor)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

Nucleotide Transformer on ClinVar Splice Acceptor variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Splice Acceptor', column 'Nucleotide Transformer'
Configuration: PhyloGPN (481 bp receptive field, about 83M parameters)Protocol: ClinVar pathogenic versus benign classification from likelihood ratios, AUROC by variant class
Dataset subset: ClinVar Splice Acceptor variants, at least one review star
0.82 AUROC (ClinVar pathogenic or likely pathogenic versus benign or likely benign, zero-shot allele log likelihood ratio, class Splice Acceptor)
unitless · higher

Uncertainty: Not reported by the source: Table 1 prints no interval or spread.

Coverage: Not reported scored / Not reported eligible

Author-reported evaluation · Source checked
Methods, coverage and source

PhyloGPN on ClinVar Splice Acceptor variants, likelihood-ratio scoring

phylogpn-20261116-protocol-clinvar-llr-auroc

Aggregation: Not reported

A Phylogenetic Approach to Genomic Language Modeling · Table 1, row 'Splice Acceptor', column 'PhyloGPN'

Source checking is not independent reproduction. Release 2026-11-21-adc15377fab2.

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
60

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

Null control

Proposed control: requires review

Training-set class prior where supervised fitting is permitted

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

Regularised classifier on simple permitted features, or protocol's conventional reference

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 2026-11-21-adc15377fab2. 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 2026-11-21-adc15377fab2
Property and statementOriginal source and locationReview and provenance

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

Sources and history

Release 2026-11-21-adc15377fab2 · Record review: source checked

2 source records and release historyDownload this release (gzip)
Technical metadata and record identifiers

Technical metadata for this record is tracked under stable ID phylogpn-20261116-protocol-clinvar-llr-auroc.

Source locator: Results, Log Likelihood Ratio Evaluation (paragraphs 1-3) and Classifying ClinVar variants; Table 1 caption; Appendix A.2

Related records

Suggest a correction