Configuration

AVI six-model ensemble (AlphaGenome Atlas v1)

Six AVI models (linear model plus linear hypernetwork plus indel embedding) whose logits are averaged; scores reported as raw logits or PHRED.

18 evaluations · 18 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.

Evaluations and results

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

Sorted by AUPRC (AVI baseline row, ClinVar column) (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	AUPRC (AVI baseline row, ClinVar column)	AUPRC (AVI baseline row, ncVar column)	AUPRC (AVI baseline row, SGE column)	AUPRC (AVI baseline row, TraitGym Complex column)	AUPRC (AVI baseline row, TraitGym Mendelian column)	AUPRC (ClinVar pathogenic vs benign SNVs, 3' UTR)	AUPRC (ClinVar pathogenic vs benign SNVs, 5' UTR)	AUPRC (ClinVar pathogenic vs benign SNVs, Intronic)	AUPRC (ClinVar pathogenic vs benign SNVs, Protein altering)	AUPRC (ClinVar pathogenic vs benign SNVs, Synonymous)	AUPRC (pooled held-out SGE screens, functionally abnormal vs normal, SNVs and indels)	AUPRC (pooled held-out SGE screens, functionally abnormal vs normal, SNVs only)	AUPRC (rare complex-trait SNVs)	AUPRC (TraitGym fine-mapped GWAS (Complex))	AUPRC (TraitGym Mendelian non-coding)	Average precision (DNM1 minigene screen, active vs inactive variants)	Recall (known P/LP variants ranked within the top 50 of each patient's variants; 105 causal variants, background filtered at gnomAD AF 0.001)	Recall (known P/LP variants ranked within the top 50 of each patient's variants; 112 causal variants, all patient variants)
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	AVI feature ablation, ClinVar column, AUPRC (AlphaGenome Atlas Table S2), ClinVar 2025-06-15 pathogenic versus benign variants, AVI evaluation set	0.978	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	ClinVar pathogenic versus benign SNVs by VEP consequence, AUPRC (AlphaGenome Atlas), ClinVar 2025-06-15 SNVs, Intronic (AlphaGenome Atlas test set)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	0.76	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	ClinVar pathogenic versus benign SNVs by VEP consequence, AUPRC (AlphaGenome Atlas), ClinVar 2025-06-15 SNVs, Protein altering (AlphaGenome Atlas test set)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	0.9	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	ClinVar pathogenic versus benign SNVs by VEP consequence, AUPRC (AlphaGenome Atlas), ClinVar 2025-06-15 SNVs, Synonymous (AlphaGenome Atlas test set)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	0.57	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	ClinVar pathogenic versus benign SNVs by VEP consequence, AUPRC (AlphaGenome Atlas), ClinVar 2025-06-15 SNVs, 3' UTR (AlphaGenome Atlas test set)	Not reported	Not reported	Not reported	Not reported	Not reported	0.5	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	ClinVar pathogenic versus benign SNVs by VEP consequence, AUPRC (AlphaGenome Atlas), ClinVar 2025-06-15 SNVs, 5' UTR (AlphaGenome Atlas test set)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	0.26	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	DNM1 minigene screen, active versus inactive variants, average precision (AlphaGenome Atlas), DNM1 intron saturation mutagenesis minigene splicing screen (AlphaGenome Atlas Data S1)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	0.708	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	GREGoR solved cases, recall of causal variants in the top 50 per patient (AlphaGenome Atlas), GREGoR solved rare disease cases, likely pathogenic and pathogenic variants (AlphaGenome Atlas)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	74.3%	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	GREGoR solved cases, recall of causal variants in the top 50 per patient (AlphaGenome Atlas), GREGoR solved rare disease cases, likely pathogenic and pathogenic variants (AlphaGenome Atlas)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	29.5%
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	Rare complex-trait SNVs versus matched controls, AUPRC (AlphaGenome Atlas), Rare variants associated with complex traits, UK Biobank and All of Us meta-analysis (AlphaGenome Atlas)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	0.69	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	SGE held-out screens pooled, SNVs and indels, AUPRC (AlphaGenome Atlas), Saturation genome editing screens of 13 genes from 14 studies (AlphaGenome Atlas)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	0.759	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	SGE held-out screens pooled, functionally abnormal versus normal SNVs, AUPRC (AlphaGenome Atlas), Saturation genome editing screens of 13 genes from 14 studies (AlphaGenome Atlas)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	0.668	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	AVI feature ablation, ncVar column, AUPRC (AlphaGenome Atlas Table S2), ncVarDB non-coding pathogenic and benign variants (as used in AlphaGenome Atlas)	Not reported	0.936	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	AVI feature ablation, SGE column, AUPRC (AlphaGenome Atlas Table S2), Saturation genome editing screens of 13 genes from 14 studies (AlphaGenome Atlas)	Not reported	Not reported	0.731	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	AVI feature ablation, TraitGym Complex column, AUPRC (AlphaGenome Atlas Table S2), TraitGym Complex traits after AVI positional filtering (AlphaGenome Atlas)	Not reported	Not reported	Not reported	0.283	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	AVI feature ablation, TraitGym Mendelian column, AUPRC (AlphaGenome Atlas Table S2), TraitGym Mendelian traits after AVI positional filtering (AlphaGenome Atlas)	Not reported	Not reported	Not reported	Not reported	0.764	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	TraitGym fine-mapped GWAS variants, AUPRC after AVI filtering (AlphaGenome Atlas), TraitGym Complex traits after AVI positional filtering (AlphaGenome Atlas)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	0.28	Not reported	Not reported	Not reported	Not reported
AVI six-model ensemble (AlphaGenome Atlas v1)Evaluation	TraitGym Mendelian non-coding variants, AUPRC after AVI filtering (AlphaGenome Atlas), TraitGym Mendelian traits after AVI positional filtering (AlphaGenome Atlas)	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	Not reported	0.76	Not reported	Not reported	Not reported

All 18 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: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: ClinVar pathogenic versus benign SNVs by VEP consequence, AUPRC (AlphaGenome Atlas) Dataset subset: ClinVar 2025-06-15 SNVs, Intronic (AlphaGenome Atlas test set)	0.76 AUPRC (ClinVar pathogenic vs benign SNVs, Intronic) unitless · higher Uncertainty: Not reported by the source: Fig. 2A draws 95% bootstrap intervals (100 bootstraps) without printing them. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source avi on ClinVar Intronic SNVs (AlphaGenome Atlas Fig. 2A) ag-atlas-20261019-protocol-clinvar-snv-by-consequence-auprc Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Results 'State-of-the-art interpretable genome-wide variant prioritization with AVI' paragraph 1 (page 5), citing Fig. 2A: 'intronic (AUPRC 0.76 vs. GPN-Star-V’s 0.44)'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: ClinVar pathogenic versus benign SNVs by VEP consequence, AUPRC (AlphaGenome Atlas) Dataset subset: ClinVar 2025-06-15 SNVs, Protein altering (AlphaGenome Atlas test set)	0.9 AUPRC (ClinVar pathogenic vs benign SNVs, Protein altering) unitless · higher Uncertainty: Not reported by the source: Fig. 2A draws 95% bootstrap intervals (100 bootstraps) without printing them. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source avi on ClinVar Protein altering SNVs (AlphaGenome Atlas Fig. 2A) ag-atlas-20261019-protocol-clinvar-snv-by-consequence-auprc Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Results 'State-of-the-art interpretable genome-wide variant prioritization with AVI' paragraph 1 (page 5), citing Fig. 2A: 'protein-altering variants (0.90 vs. GPN-Star-V’s 0.86)'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: ClinVar pathogenic versus benign SNVs by VEP consequence, AUPRC (AlphaGenome Atlas) Dataset subset: ClinVar 2025-06-15 SNVs, Synonymous (AlphaGenome Atlas test set)	0.57 AUPRC (ClinVar pathogenic vs benign SNVs, Synonymous) unitless · higher Uncertainty: Not reported by the source: Fig. 2A draws 95% bootstrap intervals (100 bootstraps) without printing them. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source avi on ClinVar Synonymous SNVs (AlphaGenome Atlas Fig. 2A) ag-atlas-20261019-protocol-clinvar-snv-by-consequence-auprc Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Results 'State-of-the-art interpretable genome-wide variant prioritization with AVI' paragraph 1 (page 5), citing Fig. 2A: 'synonymous (0.57 vs. CADD v1.7’s 0.35)'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: ClinVar pathogenic versus benign SNVs by VEP consequence, AUPRC (AlphaGenome Atlas) Dataset subset: ClinVar 2025-06-15 SNVs, 3' UTR (AlphaGenome Atlas test set)	0.5 AUPRC (ClinVar pathogenic vs benign SNVs, 3' UTR) unitless · higher Uncertainty: Not reported by the source: Fig. 2A draws 95% bootstrap intervals (100 bootstraps) without printing them. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source avi on ClinVar 3' UTR SNVs (AlphaGenome Atlas Fig. 2A) ag-atlas-20261019-protocol-clinvar-snv-by-consequence-auprc Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Results 'State-of-the-art interpretable genome-wide variant prioritization with AVI' paragraph 1 (page 5), citing Fig. 2A: '3’ UTR categories (0.50 vs. GPN-Star-M’s 0.18)'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: ClinVar pathogenic versus benign SNVs by VEP consequence, AUPRC (AlphaGenome Atlas) Dataset subset: ClinVar 2025-06-15 SNVs, 5' UTR (AlphaGenome Atlas test set)	0.26 AUPRC (ClinVar pathogenic vs benign SNVs, 5' UTR) unitless · higher Uncertainty: Not reported by the source: Fig. 2A draws 95% bootstrap intervals (100 bootstraps) without printing them. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source avi on ClinVar 5' UTR SNVs (AlphaGenome Atlas Fig. 2A) ag-atlas-20261019-protocol-clinvar-snv-by-consequence-auprc Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Results 'State-of-the-art interpretable genome-wide variant prioritization with AVI' paragraph 1 (page 5), citing Fig. 2A: 'GPN-Star-V outperformed AVI on 5’UTR ClinVar SNVs (0.27 vs. 0.26)'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: DNM1 minigene screen, active versus inactive variants, average precision (AlphaGenome Atlas) Dataset: DNM1 intron saturation mutagenesis minigene splicing screen (AlphaGenome Atlas Data S1)	0.708 Average precision (DNM1 minigene screen, active vs inactive variants) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source AVI PHRED on the DNM1 minigene splicing screen (AlphaGenome Atlas fig. S9E) ag-atlas-20261019-protocol-dnm1-active-variant-ap Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · fig. S9E legend entry 'AVI PHRED (AP=0.708)' (page 60)
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: GREGoR solved cases, recall of causal variants in the top 50 per patient (AlphaGenome Atlas) Dataset: GREGoR solved rare disease cases, likely pathogenic and pathogenic variants (AlphaGenome Atlas)	74.3% Recall (known P/LP variants ranked within the top 50 of each patient's variants; 105 causal variants, background filtered at gnomAD AF 0.001) percent · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source avi on GREGoR solved cases, af-filtered (AlphaGenome Atlas) ag-atlas-20261019-protocol-gregor-recall-top50 Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Results 'AVI prioritization helped resolve a rare disease case...' paragraph 1 (page 7)
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: GREGoR solved cases, recall of causal variants in the top 50 per patient (AlphaGenome Atlas) Dataset: GREGoR solved rare disease cases, likely pathogenic and pathogenic variants (AlphaGenome Atlas)	29.5% Recall (known P/LP variants ranked within the top 50 of each patient's variants; 112 causal variants, all patient variants) percent · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source avi on GREGoR solved cases, all-variants (AlphaGenome Atlas) ag-atlas-20261019-protocol-gregor-recall-top50 Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Results 'AVI prioritization helped resolve a rare disease case...' paragraph 1 (page 7)
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: Rare complex-trait SNVs versus matched controls, AUPRC (AlphaGenome Atlas) Dataset: Rare variants associated with complex traits, UK Biobank and All of Us meta-analysis (AlphaGenome Atlas)	0.69 AUPRC (rare complex-trait SNVs) unitless · higher Uncertainty: Not reported by the source: Fig. 2B draws 95% bootstrap intervals (100 bootstraps) without printing them. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source avi on rare complex-trait SNVs (AlphaGenome Atlas Fig. 2B) ag-atlas-20261019-protocol-rare-complex-trait-snv-auprc Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Results 'State-of-the-art interpretable genome-wide variant prioritization with AVI' paragraph 1 (page 5), citing Fig. 2B: '(AUPRC 0.69 vs. GPN-Star-M’s 0.67)'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: SGE held-out screens pooled, SNVs and indels, AUPRC (AlphaGenome Atlas) Dataset: Saturation genome editing screens of 13 genes from 14 studies (AlphaGenome Atlas)	0.759 AUPRC (pooled held-out SGE screens, functionally abnormal vs normal, SNVs and indels) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source AVI on pooled held-out SGE screens, all-variants (AlphaGenome Atlas fig. S7F) ag-atlas-20261019-protocol-sge-pooled-all-variants-auprc Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · fig. S7F (left) legend entry 'AVI (AUPRC = 0.759)' (page 57)
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: SGE held-out screens pooled, functionally abnormal versus normal SNVs, AUPRC (AlphaGenome Atlas) Dataset: Saturation genome editing screens of 13 genes from 14 studies (AlphaGenome Atlas)	0.668 AUPRC (pooled held-out SGE screens, functionally abnormal vs normal, SNVs only) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source AVI on pooled held-out SGE screens, snv (AlphaGenome Atlas fig. S7E) ag-atlas-20261019-protocol-sge-pooled-snv-auprc Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · fig. S7E (left) legend entry 'AVI (AUPRC = 0.668)' (page 57)
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: AVI feature ablation, ClinVar column, AUPRC (AlphaGenome Atlas Table S2) Dataset: ClinVar 2025-06-15 pathogenic versus benign variants, AVI evaluation set	0.978 AUPRC (AVI baseline row, ClinVar column) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source AVI AVI Model (Baseline) on ClinVar (AlphaGenome Atlas Table S2) ag-atlas-20261019-protocol-table-s2-clinvar Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Table S2 (page 70), section 'Baseline', row 'AVI Model', column 'ClinVar'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: AVI feature ablation, ncVar column, AUPRC (AlphaGenome Atlas Table S2) Dataset: ncVarDB non-coding pathogenic and benign variants (as used in AlphaGenome Atlas)	0.936 AUPRC (AVI baseline row, ncVar column) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source AVI AVI Model (Baseline) on ncVar (AlphaGenome Atlas Table S2) ag-atlas-20261019-protocol-table-s2-ncvar Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Table S2 (page 70), section 'Baseline', row 'AVI Model', column 'ncVar'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: AVI feature ablation, SGE column, AUPRC (AlphaGenome Atlas Table S2) Dataset: Saturation genome editing screens of 13 genes from 14 studies (AlphaGenome Atlas)	0.731 AUPRC (AVI baseline row, SGE column) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source AVI AVI Model (Baseline) on SGE (AlphaGenome Atlas Table S2) ag-atlas-20261019-protocol-table-s2-sge Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Table S2 (page 70), section 'Baseline', row 'AVI Model', column 'SGE'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: AVI feature ablation, TraitGym Complex column, AUPRC (AlphaGenome Atlas Table S2) Dataset subset: TraitGym Complex traits after AVI positional filtering (AlphaGenome Atlas)	0.283 AUPRC (AVI baseline row, TraitGym Complex column) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source AVI AVI Model (Baseline) on TraitGym Complex (AlphaGenome Atlas Table S2) ag-atlas-20261019-protocol-table-s2-traitgym-complex Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Table S2 (page 70), section 'Baseline', row 'AVI Model', column 'TraitGym Complex'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: AVI feature ablation, TraitGym Mendelian column, AUPRC (AlphaGenome Atlas Table S2) Dataset subset: TraitGym Mendelian traits after AVI positional filtering (AlphaGenome Atlas)	0.764 AUPRC (AVI baseline row, TraitGym Mendelian column) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source AVI AVI Model (Baseline) on TraitGym Mendelian (AlphaGenome Atlas Table S2) ag-atlas-20261019-protocol-table-s2-traitgym-mendelian Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Table S2 (page 70), section 'Baseline', row 'AVI Model', column 'TraitGym Mendelian'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: TraitGym fine-mapped GWAS variants, AUPRC after AVI filtering (AlphaGenome Atlas) Dataset subset: TraitGym Complex traits after AVI positional filtering (AlphaGenome Atlas)	0.28 AUPRC (TraitGym fine-mapped GWAS (Complex)) unitless · higher Uncertainty: Not reported by the source: Fig. 2B draws 95% bootstrap intervals (100 bootstraps) without printing them. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source avi on TraitGym fine-mapped GWAS (Complex) (AlphaGenome Atlas Fig. 2B) ag-atlas-20261019-protocol-traitgym-complex-auprc Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Results 'State-of-the-art interpretable genome-wide variant prioritization with AVI' paragraph 1 (page 5), citing Fig. 2B: 'TraitGym Complex, AUPRC 0.28 vs. 0.27'
Configuration: AVI six-model ensemble (AlphaGenome Atlas v1)	Protocol: TraitGym Mendelian non-coding variants, AUPRC after AVI filtering (AlphaGenome Atlas) Dataset subset: TraitGym Mendelian traits after AVI positional filtering (AlphaGenome Atlas)	0.76 AUPRC (TraitGym Mendelian non-coding) unitless · higher Uncertainty: Not reported by the source: Fig. 2B draws 95% bootstrap intervals (100 bootstraps) without printing them. Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source avi on TraitGym Mendelian non-coding (AlphaGenome Atlas Fig. 2B) ag-atlas-20261019-protocol-traitgym-mendelian-auprc Aggregation: Not reported AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Results 'State-of-the-art interpretable genome-wide variant prioritization with AVI' paragraph 1 (page 5), citing Fig. 2B: 'AVI ranked second to GPN-Star-M (AUPRC 0.76 vs. 0.77)'

Source checking is not independent reproduction. Release 2026-11-02-a7f03c9e55a5.

Use this configuration

How it works, versions and access

Additional methodological details, versions and access terms have not yet been extracted into a structured profile for this configuration. Consult the linked sources and evaluation records below.

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 2026-11-02-a7f03c9e55a5
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 2026-11-02-a7f03c9e55a5 · Record review: source checked

1 source record and release history

AlphaGenome Atlas: in silico mutagenesis of the entire human genome (Cheng, Taylor, Nicolaisen et al., medRxiv 2026) · Original source · medRxiv preprint version 1, posted 2026-09-20; full-text PDF with supplementary figures and tables (83 pages)

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: ag-atlas-20261019-config-avi-ensemble

areas: dna-genomes
contexts: research
method types: supervised_machine_learning
reported name: AVI
version: AlphaGenome Atlas medRxiv v1 ensemble
parameters: Ensemble of six models chosen by validation AUPRC from a sweep over 10 resampled datasets and 50 seeds; hidden layers 2x16 or 2x32, exponential or softplus weight activation, additive bias gating on or off. At inference, indel AlphaMissense is imputed as the mean SNV AlphaMissense score over the affected interval. Checkpoint selection on the average validation AUPRC of TraitGym Complex Traits, SGE SNVs and SGE indels.
training overlap: Benchmark variants whose position overlaps any AVI training or validation variant are removed for every method; comparator training sets were not filtered.
source locator: Methods 'Ensemble Models', 'Training Procedure', 'Overlap Filtering during Evaluation' (pages 22 to 25)
missing metadata: checkpoint: reason: unavailable; note: No checkpoint identifier or weights link printed.

Related records

Suggest a correction