mRNABench Sample designed MRL
Current source-reviewed mapping
Direct evidence for the stated endpoint
The matched sequence-composition and training-mean controls provide an auditable starting point for deciding whether a more complex method adds predictive value on this defined endpoint. They do not establish transfer to a different library or supply a pretrained-model comparison.
- Assessed endpoint
- Predicting target_mrl_designed on the complete 15,003-record canonical test split of the Sample designed reporter dataset.
- Evaluation protocol
- mRNABench Sample designed MRL
- Computational task
- A reviewed task relationship is not recorded for this protocol.
- Input and population constraints
- Both controls use the same full processed sequences, seed-2541 split, training labels and complete held-out test population. No validation or test labels select hyperparameters.
- Composition features are log1p length, A/C/G/T fractions and unknown fraction, fitted with train-only RidgeCV over the recorded five alpha values and no normalization. The constant control is the training-target arithmetic mean.
- MSE is the matched primary endpoint. Constant-control Pearson and Spearman are unavailable; do not replace missing values with zeros.
Limits on interpretation
- One target, one split and one execution per method; no uncertainty or seed variability. No homology-separated or compositional-generalisation claim.
- This training-only protocol differs from upstream probing and is not an aggregate mRNABench score or a reproduction of a published model result.
- The original run’s raw inputs and saved predictions are not publicly archived. The recipe obtains source data and runs the controls again, including separate protein controls; it does not rescore prior predictions. Source-data reuse terms and cross-platform reproducibility remain unresolved.
- Research baseline evidence only, with no new execution, human domain review or clinical validation.
Automated source review · 2026-09-28 · Codex research curation
Primary-source curation and separate automated cross-review checked exact evidence identities, comparator coverage, endpoint relevance and transfer limits. No new execution, human scientific review, independent replication or clinical validation.
Evaluated configurations
Each configuration below belongs to this protocol. Inspect its inputs, population and scoring conditions before comparing it with another evaluation.
Sequence composition + RidgeCV (mRNABench Sample designed MRL)
Rewire evaluation · Source checked
Predict target_mrl_designed from sequence on the complete 15,003-record canonical test split. Train-only RidgeCV differs from upstream default validation evaluation; this is not an aggregate mRNABench score.
- Population and split
- 15003/15003 · canonical test
- Inputs and adaptation
- Complete source sequence only; no assay labels during prediction · train-only fitting
- Evaluation budget
- one local CPU evaluation; no hyperparameter search outside training
- Runtime and memory
Inference and fit section: 0.955 s.
Metric evaluation section: 1.18 s.
Device: Not reported. Batch size: 32.
These timings describe the recorded sections of this run, not total runtime or a general hardware benchmark. Peak memory is not reported.
| Metric | Value | Coverage | Uncertainty and source |
|---|---|---|---|
| mse | 1.91 squared mean ribosome load · lower | 15003/15003 | Not reported Result provenance |
| pearson | 0.437 dimensionless · higher | 15003/15003 | Not reported Result provenance |
| spearman | 0.495 dimensionless · higher | 15003/15003 | Not reported Result provenance |
Uncertainty: Not estimated; one complete selected evaluation.
Evaluation methods, evidence and reproduction
Recipe: generate and evaluate predictions
This committed script regenerates the selected local evaluation with the recorded inputs and configuration. Sequence-control instructions execute all four controls; select this evaluation from their outputs. Raw prior predictions are private, so public evidence alone cannot rescore them.
Training-mean control (mRNABench Sample designed MRL)
Rewire evaluation · Source checked
Predict target_mrl_designed from sequence on the complete 15,003-record canonical test split. Train-only RidgeCV differs from upstream default validation evaluation; this is not an aggregate mRNABench score.
- Population and split
- 15003/15003 · canonical test
- Inputs and adaptation
- Complete source sequence only; no assay labels during prediction · train-only fitting
- Evaluation budget
- one local CPU evaluation; no hyperparameter search outside training
- Runtime and memory
Inference and fit section: 0.0676 s.
Metric evaluation section: 1.18 s.
Device: Not reported. Batch size: 32.
These timings describe the recorded sections of this run, not total runtime or a general hardware benchmark. Peak memory is not reported.
| Metric | Value | Coverage | Uncertainty and source |
|---|---|---|---|
| mse | 2.37 squared mean ribosome load · lower | 15003/15003 | Not reported Result provenance |
| pearson | undefined dimensionless · higher | 15003/15003 | Not reported Result provenance |
| spearman | undefined dimensionless · higher | 15003/15003 | Not reported Result provenance
|
Uncertainty: Not estimated; one complete selected evaluation.
Evaluation methods, evidence and reproduction
Recipe: generate and evaluate predictions
This committed script regenerates the selected local evaluation with the recorded inputs and configuration. Sequence-control instructions execute all four controls; select this evaluation from their outputs. Raw prior predictions are private, so public evidence alone cannot rescore them.
Open the protocol's results and comparison checks →
Original execution documentation ↗
Mapping sources and review metadata
- Sequence composition + RidgeCV: local execution report (20 September 2026) · Original source ↗
/coverage; /model_configuration; /protocol_configuration; /protocol_results; /provenance - Training-mean control: local execution report (20 September 2026) · Original source ↗
/coverage; /model_configuration; /protocol_results/metric_unavailable_reasons; /provenance - mRNABench Sample designed MRL: reproduction instructions · Original source ↗
Results: RNA evaluation paragraph; Provenance and review; Reproduce the four sequence controls
Mapping use-case-mapping-utr-translation-baselines-designed-v1 · revision 1
Initial applicability review: primary sources and separate automated cross-review support this exact protocol, its complete selected comparator group and the stated limits.
Reviewed evidence fingerprint 1034171098b5ead5f29764df1c6efaccee4aa4c24f235945e36b8536ec30caf5