Configuration

EVEE covariance probe on Evo 2 7B layer 27 (Pearce et al. 2026)

Evidence concern: excluded from comparisons

The column labelled 'NTv3' in Figure 1b and 1c is not the method the Methods describe. The manuscript repository's plotting code (scripts/figure1/_heatmap_common.py at e5e4a43) renames 'NTv3 cosine dist' to 'NTv3' and drops 'NTv3 subref probe', whereas Methods 'Baseline methods' describe NTv3 650M layer-11 embeddings with L2-regularised logistic regression under 5-fold gene-holdout cross-validation. Figure 1d's NTv3 column is the supervised subref probe. The SNV and deconfounded NTv3 values are recorded as a zero-shot embedding distance.Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4 · Figure 1b and 1c, column 'NTv3' (page 3); Methods 'Baseline methods', NTv3 (page 8); repository scripts/figure1/_heatmap_common.py RENAME_MAP and DROP_METHODS

AlphaMissense's 'Overall' cell in Figure 1b and 1c is its missense AUROC. The plotting code blanks AlphaMissense outside missense and copies the Missense value into the Overall row (scripts/figure1/_heatmap_common.py, am_ov_mask), so it is computed on a different, missense-only population from the other methods' Overall cells.Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4 · Figure 1b and 1c, row 'Overall', column 'AlphaMissense' (page 3); repository scripts/figure1/_heatmap_common.py prepare()

Covariance probe on alternate-minus-reference Evo 2 7B layer-27 embeddings, 65 kb window, top 256 most divergent positions per strand, trained on ClinVar SNVs with an 80/20 gene-level split.

3 evaluations · 3 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

3 evaluations · 3 results. Different protocols are not a single leaderboard.

Sorted by Accelerator time (Evo 2 7B embedding extraction for 4,252,870 ClinVar variants) (lower 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	Accelerator time (Evo 2 7B embedding extraction for 4,252,870 ClinVar variants)	AUROC (ClinVar indels, all strata (Overall row); zero-shot transfer from SNV training)	AUROC (Deconfounded ClinVar SNVs, all consequence types (Overall row))
EVEE covariance probe on Evo 2 7B layer 27 (Pearce et al. 2026)Evaluation	EVEE Evo 2 embedding extraction compute (Pearce et al. 2026), EVEE full ClinVar variant set (4,252,870 variants)	∼20,000	Not reported	Not reported
EVEE covariance probe on Evo 2 7B layer 27 (Pearce et al. 2026)Evaluation	EVEE deconfounded ClinVar SNV pathogenicity AUROC (Pearce et al. 2026 Figure 1c), EVEE deconfounded ClinVar SNV benchmark (158,616 variants)	Not reported	Not reported	0.977
EVEE covariance probe on Evo 2 7B layer 27 (Pearce et al. 2026)Evaluation	EVEE ClinVar indel pathogenicity AUROC (Pearce et al. 2026 Figure 1d), EVEE ClinVar indel benchmark (73,961 indels)	Not reported	0.991	Not reported

All 3 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: EVEE covariance probe on Evo 2 7B layer 27 (Pearce et al. 2026)	Protocol: EVEE deconfounded ClinVar SNV pathogenicity AUROC (Pearce et al. 2026 Figure 1c) Dataset: EVEE deconfounded ClinVar SNV benchmark (158,616 variants)	0.977 AUROC (Deconfounded ClinVar SNVs, all consequence types (Overall row)) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Evidence concern: excluded from comparisons The column labelled 'NTv3' in Figure 1b and 1c is not the method the Methods describe. The manuscript repository's plotting code (scripts/figure1/_heatmap_common.py at e5e4a43) renames 'NTv3 cosine dist' to 'NTv3' and drops 'NTv3 subref probe', whereas Methods 'Baseline methods' describe NTv3 650M layer-11 embeddings with L2-regularised logistic regression under 5-fold gene-holdout cross-validation. Figure 1d's NTv3 column is the supervised subref probe. The SNV and deconfounded NTv3 values are recorded as a zero-shot embedding distance.Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4 · Figure 1b and 1c, column 'NTv3' (page 3); Methods 'Baseline methods', NTv3 (page 8); repository scripts/figure1/_heatmap_common.py RENAME_MAP and DROP_METHODS AlphaMissense's 'Overall' cell in Figure 1b and 1c is its missense AUROC. The plotting code blanks AlphaMissense outside missense and copies the Missense value into the Overall row (scripts/figure1/_heatmap_common.py, am_ov_mask), so it is computed on a different, missense-only population from the other methods' Overall cells.Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4 · Figure 1b and 1c, row 'Overall', column 'AlphaMissense' (page 3); repository scripts/figure1/_heatmap_common.py prepare() Methods, coverage and source Evo2 Covariance on the EVEE deconfounded ClinVar SNV benchmark (Pearce et al. 2026) gfm-guide-20261014-protocol-evee-deconfounded-snv-auroc Aggregation: Not reported Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4; EVEE manuscript repository: figures/figure1/fig1c_deconf_heatmap.pdf; EVEE manuscript repository: scripts/figure1/_heatmap_common.py · Figure 1c (page 3, raster; matched to repository figures/figure1/fig1c_deconf_heatmap.pdf text layer), row 'Overall (n=158,616, 27.8%P)', column 'Evo2 Covariance'
Configuration: EVEE covariance probe on Evo 2 7B layer 27 (Pearce et al. 2026)	Protocol: EVEE Evo 2 embedding extraction compute (Pearce et al. 2026) Dataset: EVEE full ClinVar variant set (4,252,870 variants)	∼20,000 Accelerator time (Evo 2 7B embedding extraction for 4,252,870 ClinVar variants) hour · lower Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Evidence concern: excluded from comparisons The column labelled 'NTv3' in Figure 1b and 1c is not the method the Methods describe. The manuscript repository's plotting code (scripts/figure1/_heatmap_common.py at e5e4a43) renames 'NTv3 cosine dist' to 'NTv3' and drops 'NTv3 subref probe', whereas Methods 'Baseline methods' describe NTv3 650M layer-11 embeddings with L2-regularised logistic regression under 5-fold gene-holdout cross-validation. Figure 1d's NTv3 column is the supervised subref probe. The SNV and deconfounded NTv3 values are recorded as a zero-shot embedding distance.Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4 · Figure 1b and 1c, column 'NTv3' (page 3); Methods 'Baseline methods', NTv3 (page 8); repository scripts/figure1/_heatmap_common.py RENAME_MAP and DROP_METHODS AlphaMissense's 'Overall' cell in Figure 1b and 1c is its missense AUROC. The plotting code blanks AlphaMissense outside missense and copies the Missense value into the Overall row (scripts/figure1/_heatmap_common.py, am_ov_mask), so it is computed on a different, missense-only population from the other methods' Overall cells.Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4 · Figure 1b and 1c, row 'Overall', column 'AlphaMissense' (page 3); repository scripts/figure1/_heatmap_common.py prepare() Methods, coverage and source EVEE Evo 2 7B embedding extraction compute (Pearce et al. 2026) gfm-guide-20261014-protocol-evee-embedding-extraction-compute Aggregation: Not reported Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4 · Methods 'Harvesting', page 8, line 461: 'Extraction used ∼20,000 H100-hours'
Configuration: EVEE covariance probe on Evo 2 7B layer 27 (Pearce et al. 2026)	Protocol: EVEE ClinVar indel pathogenicity AUROC (Pearce et al. 2026 Figure 1d) Dataset: EVEE ClinVar indel benchmark (73,961 indels)	0.991 AUROC (ClinVar indels, all strata (Overall row); zero-shot transfer from SNV training) unitless · higher Uncertainty: Not reported by the source Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Evidence concern: excluded from comparisons The column labelled 'NTv3' in Figure 1b and 1c is not the method the Methods describe. The manuscript repository's plotting code (scripts/figure1/_heatmap_common.py at e5e4a43) renames 'NTv3 cosine dist' to 'NTv3' and drops 'NTv3 subref probe', whereas Methods 'Baseline methods' describe NTv3 650M layer-11 embeddings with L2-regularised logistic regression under 5-fold gene-holdout cross-validation. Figure 1d's NTv3 column is the supervised subref probe. The SNV and deconfounded NTv3 values are recorded as a zero-shot embedding distance.Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4 · Figure 1b and 1c, column 'NTv3' (page 3); Methods 'Baseline methods', NTv3 (page 8); repository scripts/figure1/_heatmap_common.py RENAME_MAP and DROP_METHODS AlphaMissense's 'Overall' cell in Figure 1b and 1c is its missense AUROC. The plotting code blanks AlphaMissense outside missense and copies the Missense value into the Overall row (scripts/figure1/_heatmap_common.py, am_ov_mask), so it is computed on a different, missense-only population from the other methods' Overall cells.Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4 · Figure 1b and 1c, row 'Overall', column 'AlphaMissense' (page 3); repository scripts/figure1/_heatmap_common.py prepare() Methods, coverage and source Evo2 Covariance on the EVEE ClinVar indel benchmark (Pearce et al. 2026) gfm-guide-20261014-protocol-evee-clinvar-indel-auroc Aggregation: Not reported Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4; EVEE manuscript repository: figures/figure1/fig1d_indel_heatmap.pdf; EVEE manuscript repository: scripts/figure1/fig1d_indel_heatmap.py · Figure 1d (page 3, raster; matched to repository figures/figure1/fig1d_indel_heatmap.pdf text layer), row 'Overall (n=73,961, 65.1%P)', column 'Evo2 Covariance'

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

Use this configuration

How it works, versions and access

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

1 source record and release history

Interpretable variant effect prediction from genomic foundation model representations (EVEE), bioRxiv v4 · Original source · bioRxiv v4, posted 2026-07-11 (v1 posted 2026-04-11)

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: gfm-guide-20261014-config-evee-evo2-covariance

areas: dna-genomes
contexts: research
method types: foundation_model; supervised_machine_learning
source locator: Methods 'Baseline methods' and 'Covariance probe' (page 8); Figure 1 column labels
reported name: Evo2 Covariance
version: Deployed probe as described in bioRxiv v4
checkpoint: Evo 2 7B, layer 27

Related records

configuration of: EVEE
uses model: Evo 2
system: Evo2 Covariance on the EVEE deconfounded ClinVar SNV benchmark (Pearce et al. 2026)
system: EVEE Evo 2 7B embedding extraction compute (Pearce et al. 2026)
system: Evo2 Covariance on the EVEE ClinVar indel benchmark (Pearce et al. 2026)

Suggest a correction