AUROC (task: pathogenic vs common snp; outer test group 3 only (chromosomes 1, 4, 7, 8, 10, 13, 15), not averaged) of Caduceus-Ph, long sequence (pathogenic/common SNP classification) on Feng pathogenic/common SNP long-window dataset
0.625 AUROC (task: pathogenic vs common snp; outer test group 3 only (chromosomes 1, 4, 7, 8, 10, 13, 15), not averaged)
Methods
- Tested configuration
- Caduceus-Ph, long sequence (pathogenic/common SNP classification)
- Protocol
- Feng pathogenic-versus-common SNP classification
- Dataset
- Feng pathogenic/common SNP long-window dataset
- Procedure
- Table 5 set-up, AUC reported separately for each outer chromosome test group (Supplementary Data 4); frozen embeddings + random forest classifier
- Evaluation
- Caduceus-Ph, long sequence Pathogenic variants vs common per-test-group evaluation
- Coverage
- Not reported scored / Not reported eligible
- Uncertainty
- Not reported by the source: Supplementary Data 4 prints one AUC per test group, with no interval.
- Evidence
- Independent external evaluation · source checkedBenchmarking DNA foundation models for genomic and genetic tasks: Supplementary Data 1-6 · Supplementary Data 4 (workbook sheet 'Data 4'), block 'Pathogenic variants vs common', row 'Caduceus-Ph, long sequence', column 'Test Group 3 AUC' (cell D4). Test group 3 is chromosomes 1, 4, 7, 8, 10, 13, 15 (Methods, 'Variant effect quantification benchmark').
A source-checked result verifies the numerical transcription, not every model or protocol detail. Evaluation metadata: source checked. Source checked does not mean independently reproduced.
Reproduction
- Split
- Three disjoint outer chromosome test groups (3/6/9/12/16/18/19/21; 2/5/11/14/17/20/22/X; 1/4/7/8/10/13/15), each held out in turn with fourfold inner-chromosome cross-validation for hyperparameter tuning.
- Adaptation
- Frozen pretrained representation plus a supervised random-forest classifier head fitted for this task
- Scoring implementation
- Not reported
No execution recipe has been verified for this exact configuration and evaluation. A benchmark's general instructions may use different inputs, splits or model settings.
Reproducing this published result requires matching its model configuration, data, split and scorer. Source checking or a successful smoke test does not establish score reproduction.
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.
1 evidence row matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Reported result 0.6254 Individual claims | Benchmarking DNA foundation models for genomic and genetic tasks: Supplementary Data 1-6 Supplementary Data 4 (workbook sheet 'Data 4'), block 'Pathogenic variants vs common', row 'Caduceus-Ph, long sequence', column 'Test Group 3 AUC' (cell D4). Test group 3 is chromosomes 1, 4, 7, 8, 10, 13, 15 (Methods, 'Variant effect quantification benchmark'). Version: Nature Communications 16, 10780 (2025); Supplementary Data 1-6 workbook (MOESM3, .xlsm) | source checked ["source-hash-verification","deterministic-table-parse","independent-cell-check"] · 2026-10-14T15:00:00Z independent paper Audit detailsExtracted by extract/extract_feng.py (deterministic parse). Independent review: value, metric and identity match the source. The workbook stores the cell as a binary double in General number format; printed_value is the value as displayed. This per-group AUC is one of the three test-set values whose mean is the Table 5 cell already stored as amp-feng-20261007-result-variant-caduceus-ph-long-sequence-auc on evaluation amp-feng-20261007-eval-variant-caduceus-ph-long-sequence. It sits on its own evaluation because its aggregation differs. Field: Source artifact SHA-256: Hash scope: Hash scope not separately documented; inspect source record Extraction artifact SHA-256: |
Sources and history
Release 2026-11-02-a7f03c9e55a5 · Record review: source checked
2 source records and release history
- Benchmarking DNA foundation models for genomic and genetic tasks: Supplementary Data 1-6 · Original source · Nature Communications 16, 10780 (2025); Supplementary Data 1-6 workbook (MOESM3, .xlsm)
- Benchmarking DNA foundation models for genomic and genetic tasks · Original source · PMC12663285.1
Technical metadata and extraction receipts
Stable ID: feng-dna-fm-20261014-result-sd4-pathogenic-common-caduceus-ph-long-sequence-group-3-auc
- metric
- auroc
- metric direction
- higher
- unit
- unitless
- printed value
- 0.6254
- numeric value
- 0.6254
- metric qualifier
- task: pathogenic vs common snp; outer test group 3 only (chromosomes 1, 4, 7, 8, 10, 13, 15), not averaged
- source locator
- Supplementary Data 4 (workbook sheet 'Data 4'), block 'Pathogenic variants vs common', row 'Caduceus-Ph, long sequence', column 'Test Group 3 AUC' (cell D4). Test group 3 is chromosomes 1, 4, 7, 8, 10, 13, 15 (Methods, 'Variant effect quantification benchmark').
- missing metadata
- uncertainty: reason: unreported; note: Supplementary Data 4 prints one AUC per test group, with no interval.
- review
- method: source-hash-verification; deterministic-table-parse; independent-cell-check; method note: Re-downloaded the Supplementary Data workbook and matched the pinned SHA-256. Read sheet 'Data 4' with a separate parser (review/signoff.py) and checked the cell, printed and numeric value, qualifier, and that the result's evaluation has the same configuration, protocol and dataset as the stored Table 5/6 evaluation, whose printed mean is within rounding of the mean of the three groups.; reviewer: claude; reviewer note: Independent Claude (Opus 5.5) review agent for PR #157, separate from the extraction agent for issue #142 that built this batch. Automated review only; no human review claimed. Transcription, not reproduction.; reviewed at: 2026-10-14T15:00:00Z; artifact sha256: 4804c801e0e054eca5f776617c18a6a0cb31d6ce3d2a7a52f70e4dda5b3abd7a; retrieval url: https://static-content.springer.com/esm/art%3A10.1038%2Fs41467-025-65823-8/MediaObjects/41467_2025_65823_MOESM3_ESM.xlsm; note: Extracted by extract/extract_feng.py (deterministic parse). Independent review: value, metric and identity match the source. The workbook stores the cell as a binary double in General number format; printed_value is the value as displayed. This per-group AUC is one of the three test-set values whose mean is the Table 5 cell already stored as amp-feng-20261007-result-variant-caduceus-ph-long-sequence-auc on evaluation amp-feng-20261007-eval-variant-caduceus-ph-long-sequence. It sits on its own evaluation because its aggregation differs.