mRNABench variant probes VEP: Variant effect prediction
Variant effect prediction. Scored with AUPRC on mRNABench VEP (mRNABench split). Frozen transcript representations with task-specific linear probes, or named supervised/naive controls. Homology splits where possible; random splits for MRL-MPRA, MRL-HL-Pair and VEP. Reported mean over ten splits; Appendix C enumerates nine seeds. Table captions say ten random splits, while Methods specify homology splits where possible. This source ambiguity and the nine explicitly listed seeds are retained; exact split manifests remain unextracted. Appendix C uses micro-averaged multilabel metrics.
Overview
Variant effect prediction. Scored with AUPRC on mRNABench VEP (mRNABench split). Frozen transcript representations with task-specific linear probes, or named supervised/naive controls. Homology splits where possible; random splits for MRL-MPRA, MRL-HL-Pair and VEP. Reported mean over ten splits; Appendix C enumerates nine seeds. Table captions say ten random splits, while Methods specify homology splits where possible. This source ambiguity and the nine explicitly listed seeds are retained; exact split manifests remain unextracted. Appendix C uses micro-averaged multilabel metrics.
Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.
Results
Each comparison retains its reviewed evaluation scope, dataset and metric. Results are shown without a pooled ranking.
mRNABench variant probes VEP: Variant effect prediction
auprc (fraction) · Higher values are better.
Every method mRNABench variant probes reports on Variant effect prediction, scored with AUPRC on mRNABench VEP (mRNABench split).
mRNABench variant probes VEP: Variant effect prediction · mRNABench VEP (mRNABench split)
Evidence origin: Author-reported evaluation. Numerical source review does not establish independent reproduction.
mRNABench: A curated benchmark for mature mRNA property and function prediction · Table 5 (XML T2), VEP column; Methods S17/S18; Appendix C (APP3)These are configuration-level details of the experiment set summarized by the paper’s Table 2; they are not new independent experiments.
All comparison limitations (5)
- These are configuration-level details of the experiment set summarized by the paper’s Table 2; they are not new independent experiments.
- Tables report means over ten splits and 95% confidence intervals. Appendix C lists only nine seeds; the missing seed is not inferred.
- Two localization columns are excluded because Table 2 and Table 5 disagree about their metric.
- Supervised CNN and naive feature controls are retained; not every method is a pretrained model.
- Table captions say ten random splits, while Methods specify homology splits where possible. This source ambiguity and the nine explicitly listed seeds are retained; exact split manifests remain unextracted.
Automated source review: 2026-09-23.
No unavailable values; missing scores remain labelled and are never plotted as zero.
Showing 12 of 50 matching rows.
- supervised-cnn0.332 ± 0.023
- orthrus-mlm0.323 ± 0.032
- evo2–7b-base0.321 ± 0.032
- evo2–7b0.320 ± 0.031
- evo2–1b-base0.319 ± 0.028
- nt-2.5b-multi-species0.318 ± 0.031
- aido-rna-1b600m0.316 ± 0.032
- rinalmo0.316 ± 0.027
- aido-rna-650m-cds0.312 ± 0.036
- utrbert-6mer0.312 ± 0.029
- aido-rna-650m0.309 ± 0.031
- mrna-fm0.307 ± 0.031
Dots show point estimates. Whiskers show only explicitly defined uncertainty (standard deviation, standard error or a labelled interval); their definitions remain in Table. Unresolved uncertainty is not plotted. Differences do not establish statistical significance.
Methods and evaluation design
Procedure, tasks and evaluated configurations
Evaluation design
Benchmarks bring together tasks and protocols. A task describes the biological question; a protocol defines a particular test.
Benchmarks
These source-backed links do not make different protocols or scores interchangeable.
Recorded evaluations
Each evaluation records what was tested and under which conditions.
- aido-rna-1b600m-cds on mRNABench variant probes VEP: Variant effect prediction
- aido-rna-1b600m on mRNABench variant probes VEP: Variant effect prediction
- aido-rna-650m-cds on mRNABench variant probes VEP: Variant effect prediction
- aido-rna-650m on mRNABench variant probes VEP: Variant effect prediction
- dnabert-s on mRNABench variant probes VEP: Variant effect prediction
- dnabert2 on mRNABench variant probes VEP: Variant effect prediction
- ernierna-ss on mRNABench variant probes VEP: Variant effect prediction
- ernierna on mRNABench variant probes VEP: Variant effect prediction
- evo-1–131k-base on mRNABench variant probes VEP: Variant effect prediction
- evo-1.5–8k-base on mRNABench variant probes VEP: Variant effect prediction
- evo-1–8k-base on mRNABench variant probes VEP: Variant effect prediction
- evo2–1b-base on mRNABench variant probes VEP: Variant effect prediction
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
- 50
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 mean 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
Simple features with train-only ridge regression
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 · Model evaluation matrix · Source table · Release and checksums
Coverage is derived from release 2026-09-23-2b89723c6dd9. Source citations describe the original records; they do not validate an unreviewed baseline proposal. No results have been generated by this audit.
Run instructions
No runnable recipe has been reviewed for this protocol. Dataset access, model requirements, licences and compute requirements must be checked against its sources before execution.
Strengths, limitations and unresolved questions
Evidence
Source checking verifies the cited claim or transcription. It does not establish independent reproduction.
Evidence table
Inspect claims, sources and review details
Trace each statement to its source and review. A context-only reference supports the record generally; it does not verify an individual field. Source checking does not reproduce an experiment.
One row per statement and cited source. Multiple citations are not independent evaluations. Shared locators are labelled explicitly.
1 evidence row matching the loaded filters
| Property and statement | Original source and location | Review and provenance |
|---|---|---|
| Relationship: part of discovery-benchmark-mrnabench Individual claims | mRNABench: A curated benchmark for mature mRNA property and function prediction Table 5 (XML T2), VEP column; Methods S17/S18; Appendix C (APP3) Version: preprint archived 2025-07-08 | source checked automated source review · 2026-09-23 Audit detailsPrimary-source transcription with no human sign-off and no independent reproduction. Field: Claim: mrnabench-variants-2025-association-vep Source artifact SHA-256: Hash scope: Exact retrieved primary paper artifact bytes. |
Sources and history
View linked audit checks and correction history
Release 2026-09-23-2b89723c6dd9 · Record review: source checked
1 source records and release history
- mRNABench: A curated benchmark for mature mRNA property and function prediction · Original source · preprint archived 2025-07-08
Technical metadata and extraction receipts
Stable ID: mrnabench-variants-2025-task-vep
- areas
- rna-transcriptomics
- tasks
- Variant effect prediction
- metric
- AUPRC
- metric direction
- higher
- dataset
- mRNABench VEP (mRNABench split)
- protocol
- Frozen transcript representations with task-specific linear probes, or named supervised/naive controls. Homology splits where possible; random splits for MRL-MPRA, MRL-HL-Pair and VEP. Reported mean over ten splits; Appendix C enumerates nine seeds. Table captions say ten random splits, while Methods specify homology splits where possible. This source ambiguity and the nine explicitly listed seeds are retained; exact split manifests remain unextracted. Appendix C uses micro-averaged multilabel metrics.
- source locator
- Table 5 (XML T2), VEP column; Methods S17/S18; Appendix C (APP3)
- comparison panels
- id: mrnabench-variants-2025-panel-vep; title: mRNABench variant probes VEP: Variant effect prediction; protocol id: mrnabench-variants-2025-task-vep; dataset id: mrnabench-dataset-mrnabench-vep; metric: auprc; unit: fraction; direction: higher; result ids: mrnabench-variants-2025-result-aido-rna-1b600m-vep-auprc; mrnabench-variants-2025-result-aido-rna-1b600m-cds-vep-auprc; mrnabench-variants-2025-result-aido-rna-650m-vep-auprc; mrnabench-variants-2025-result-aido-rna-650m-cds-vep-auprc; mrnabench-variants-2025-result-dnabert-s-vep-auprc; mrnabench-variants-2025-result-dnabert2-vep-auprc; mrnabench-variants-2025-result-ernierna-vep-auprc; mrnabench-variants-2025-result-ernierna-ss-vep-auprc; mrnabench-variants-2025-result-evo-1-131k-base-vep-auprc; mrnabench-variants-2025-result-evo-1-8k-base-vep-auprc; mrnabench-variants-2025-result-evo-1-5-8k-base-vep-auprc; mrnabench-variants-2025-result-evo2-1b-base-vep-auprc; mrnabench-variants-2025-result-evo2-7b-vep-auprc; mrnabench-variants-2025-result-evo2-7b-base-vep-auprc; mrnabench-variants-2025-result-helix-mrna-vep-auprc; mrnabench-variants-2025-result-hyenadna-large-1m-vep-auprc; mrnabench-variants-2025-result-hyenadna-medium-160k-vep-auprc; mrnabench-variants-2025-result-hyenadna-medium-450k-vep-auprc; mrnabench-variants-2025-result-hyenadna-small-32k-vep-auprc; mrnabench-variants-2025-result-hyenadna-tiny-16k-d128-vep-auprc; mrnabench-variants-2025-result-mrna-fm-vep-auprc; mrnabench-variants-2025-result-naive-4-vep-auprc; mrnabench-variants-2025-result-naive-6-vep-auprc; mrnabench-variants-2025-result-naive-mamba-vep-auprc; mrnabench-variants-2025-result-nt-2-5b-1000g-vep-auprc; mrnabench-variants-2025-result-nt-2-5b-multi-species-vep-auprc; mrnabench-variants-2025-result-nt-500m-1000g-vep-auprc; mrnabench-variants-2025-result-nt-500m-human-ref-vep-auprc; mrnabench-variants-2025-result-nt-v2-100m-multi-species-vep-auprc; mrnabench-variants-2025-result-nt-v2-250m-multi-species-vep-auprc; mrnabench-variants-2025-result-nt-v2-500m-multi-species-vep-auprc; mrnabench-variants-2025-result-nt-v2-50m-multi-species-vep-auprc; mrnabench-variants-2025-result-orthrus-base-4-vep-auprc; mrnabench-variants-2025-result-orthrus-large-6-vep-auprc; mrnabench-variants-2025-result-orthrus-mlm-vep-auprc; mrnabench-variants-2025-result-rinalmo-vep-auprc; mrnabench-variants-2025-result-rna-fm-vep-auprc; mrnabench-variants-2025-result-rnabert-vep-auprc; mrnabench-variants-2025-result-rnaernie-vep-auprc; mrnabench-variants-2025-result-rnamsm-vep-auprc; mrnabench-variants-2025-result-splicebert-h-510nt-vep-auprc; mrnabench-variants-2025-result-splicebert-v-1024nt-vep-auprc; mrnabench-variants-2025-result-splicebert-v-510nt-vep-auprc; mrnabench-variants-2025-result-supervised-cnn-vep-auprc; mrnabench-variants-2025-result-utrbert-3mer-vep-auprc; mrnabench-variants-2025-result-utrbert-4mer-vep-auprc; mrnabench-variants-2025-result-utrbert-5mer-vep-auprc; mrnabench-variants-2025-result-utrbert-6mer-vep-auprc; mrnabench-variants-2025-result-utrlm-mrl-vep-auprc; mrnabench-variants-2025-result-utrlm-te-el-vep-auprc; source ids: expansion-p3-mrnabench-2025; source locator: Table 5 (XML T2), VEP column; Methods S17/S18; Appendix C (APP3); context: Every method mRNABench variant probes reports on Variant effect prediction, scored with AUPRC on mRNABench VEP (mRNABench split).; caveats: These are configuration-level details of the experiment set summarized by the paper’s Table 2; they are not new independent experiments.; Tables report means over ten splits and 95% confidence intervals. Appendix C lists only nine seeds; the missing seed is not inferred.; Two localization columns are excluded because Table 2 and Table 5 disagree about their metric.; Supervised CNN and naive feature controls are retained; not every method is a pretrained model.; Table captions say ten random splits, while Methods specify homology splits where possible. This source ambiguity and the nine explicitly listed seeds are retained; exact split manifests remain unextracted.; review: method: automated_source_review; date: 2026-09-23
- entity level
- protocol
Related records
- part of: mRNABench
- subject: mRNABench variant probes VEP: part of discovery-benchmark-mrnabench
- benchmark: aido-rna-1b600m-cds on mRNABench variant probes VEP: Variant effect prediction
- benchmark: aido-rna-1b600m on mRNABench variant probes VEP: Variant effect prediction
- benchmark: aido-rna-650m-cds on mRNABench variant probes VEP: Variant effect prediction
- benchmark: aido-rna-650m on mRNABench variant probes VEP: Variant effect prediction
- benchmark: dnabert-s on mRNABench variant probes VEP: Variant effect prediction
- benchmark: dnabert2 on mRNABench variant probes VEP: Variant effect prediction
- benchmark: ernierna-ss on mRNABench variant probes VEP: Variant effect prediction
- benchmark: ernierna on mRNABench variant probes VEP: Variant effect prediction
- benchmark: evo-1–131k-base on mRNABench variant probes VEP: Variant effect prediction
- benchmark: evo-1.5–8k-base on mRNABench variant probes VEP: Variant effect prediction
- benchmark: evo-1–8k-base on mRNABench variant probes VEP: Variant effect prediction
- benchmark: evo2–1b-base on mRNABench variant probes VEP: Variant effect prediction
- benchmark: evo2–7b-base on mRNABench variant probes VEP: Variant effect prediction
- benchmark: evo2–7b on mRNABench variant probes VEP: Variant effect prediction
- benchmark: helix-mrna on mRNABench variant probes VEP: Variant effect prediction
- benchmark: hyenadna-large-1m on mRNABench variant probes VEP: Variant effect prediction
- benchmark: hyenadna-medium-160k on mRNABench variant probes VEP: Variant effect prediction
- benchmark: hyenadna-medium-450k on mRNABench variant probes VEP: Variant effect prediction
- benchmark: hyenadna-small-32k on mRNABench variant probes VEP: Variant effect prediction
- benchmark: hyenadna-tiny-16k-d128 on mRNABench variant probes VEP: Variant effect prediction
- benchmark: mrna-fm on mRNABench variant probes VEP: Variant effect prediction
- benchmark: naive-4 on mRNABench variant probes VEP: Variant effect prediction
- benchmark: naive-6 on mRNABench variant probes VEP: Variant effect prediction
- benchmark: naive-mamba on mRNABench variant probes VEP: Variant effect prediction
- benchmark: nt-2.5b-1000g on mRNABench variant probes VEP: Variant effect prediction
- benchmark: nt-2.5b-multi-species on mRNABench variant probes VEP: Variant effect prediction
- benchmark: nt-500m-1000g on mRNABench variant probes VEP: Variant effect prediction
- benchmark: nt-500m-human-ref on mRNABench variant probes VEP: Variant effect prediction
- benchmark: nt-v2–100m-multi-species on mRNABench variant probes VEP: Variant effect prediction
- benchmark: nt-v2–250m-multi-species on mRNABench variant probes VEP: Variant effect prediction
- benchmark: nt-v2–500m-multi-species on mRNABench variant probes VEP: Variant effect prediction
- benchmark: nt-v2–50m-multi-species on mRNABench variant probes VEP: Variant effect prediction
- benchmark: orthrus-base-4 on mRNABench variant probes VEP: Variant effect prediction
- benchmark: orthrus-large-6 on mRNABench variant probes VEP: Variant effect prediction
- benchmark: orthrus-mlm on mRNABench variant probes VEP: Variant effect prediction
- benchmark: rinalmo on mRNABench variant probes VEP: Variant effect prediction
- benchmark: rna-fm on mRNABench variant probes VEP: Variant effect prediction
- benchmark: rnabert on mRNABench variant probes VEP: Variant effect prediction
- benchmark: rnaernie on mRNABench variant probes VEP: Variant effect prediction
- benchmark: rnamsm on mRNABench variant probes VEP: Variant effect prediction
- benchmark: splicebert-h-510nt on mRNABench variant probes VEP: Variant effect prediction
- benchmark: splicebert-v-1024nt on mRNABench variant probes VEP: Variant effect prediction
- benchmark: splicebert-v-510nt on mRNABench variant probes VEP: Variant effect prediction
- benchmark: supervised-cnn on mRNABench variant probes VEP: Variant effect prediction
- benchmark: utrbert-3mer on mRNABench variant probes VEP: Variant effect prediction
- benchmark: utrbert-4mer on mRNABench variant probes VEP: Variant effect prediction
- benchmark: utrbert-5mer on mRNABench variant probes VEP: Variant effect prediction
- benchmark: utrbert-6mer on mRNABench variant probes VEP: Variant effect prediction
- benchmark: utrlm-mrl on mRNABench variant probes VEP: Variant effect prediction
- benchmark: utrlm-te-el on mRNABench variant probes VEP: Variant effect prediction