rewirebio.iobenchmarks
Pipeline

NT-v2

Embeds mature mRNA with a Nucleotide Transformer model reported as NT-v2, from a family of transformer DNA foundation models (50 million to 2.5 billion parameters, integrating information from 3,202 human genomes and 850 genomes of other species), for the mRNABench property prediction tasks.

7 evaluations · 7 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.

Model Capabilities & Hardware Requirements

Deployment facts recorded for NT-v2. Fields the record does not carry are marked not recorded.

ParametersNot recorded
Minimum VRAMNot recorded
Inference LatencyNot recorded
Context WindowNot recorded
AccessNot recorded
LicenseNot recorded

Evaluations and results

7 evaluations · 7 results. Different protocols are not a single leaderboard.

Sorted by AUPRC (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.

All 7 result rows with coverage, uncertainty and sources
Filter evaluations

Applied filters: All linked evaluations

Exact evaluated configurations and original reported results
Tested configurationProtocol and datasetFindingEvidence and details
Pipeline: NT-v2Protocol: MIMIC mRNABench probes eCLIP: eCLIP prediction
Dataset subset: mRNABench eCLIP as reported in MIMIC Table S11 (MIMIC mRNABench probes split)
0.391 AUPRC
fraction · higher

Uncertainty: Not yet extracted

Coverage: Not reported scored / Not reported eligible

Result quoted from another source · Source checked
Methods, coverage and source

NT-v2 on MIMIC mRNABench probes eCLIP: eCLIP prediction

MIMIC uses concatenated register tokens and mean-pooled RNA-track representations with supervised probes. Nucleotide/codon/amino-acid conditioning is selected per task on validation data; predictions are averaged across selected modality subsets. Comparator scores are quoted from prior mRNABench work.

Aggregation: Not reported

MIMIC v1: Table S11 and Appendix D.3 · Table S11 (HTML A4.T11), data row 9 (NT-v2 [ 121 ]), column 2 (eCLIP)
Pipeline: NT-v2Protocol: MIMIC mRNABench probes GO: GO prediction
Dataset subset: mRNABench GO as reported in MIMIC Table S11 (MIMIC mRNABench probes split)
0.35 AUPRC
fraction · higher

Uncertainty: Not yet extracted

Coverage: Not reported scored / Not reported eligible

Result quoted from another source · Source checked
Methods, coverage and source

NT-v2 on MIMIC mRNABench probes GO: GO prediction

MIMIC uses concatenated register tokens and mean-pooled RNA-track representations with supervised probes. Nucleotide/codon/amino-acid conditioning is selected per task on validation data; predictions are averaged across selected modality subsets. Comparator scores are quoted from prior mRNABench work.

Aggregation: Not reported

MIMIC v1: Table S11 and Appendix D.3 · Table S11 (HTML A4.T11), data row 9 (NT-v2 [ 121 ]), column 3 (GO)
Pipeline: NT-v2Protocol: MIMIC mRNABench probes MRL (LBKWK): MRL (LBKWK) prediction
Dataset subset: mRNABench MRL (LBKWK) as reported in MIMIC Table S11 (MIMIC mRNABench probes split)
0.512 Pearson
unitless · higher

Uncertainty: Not yet extracted

Coverage: Not reported scored / Not reported eligible

Result quoted from another source · Source checked
Methods, coverage and source

NT-v2 on MIMIC mRNABench probes MRL (LBKWK): MRL (LBKWK) prediction

MIMIC uses concatenated register tokens and mean-pooled RNA-track representations with supervised probes. Nucleotide/codon/amino-acid conditioning is selected per task on validation data; predictions are averaged across selected modality subsets. Comparator scores are quoted from prior mRNABench work.

Aggregation: Not reported

MIMIC v1: Table S11 and Appendix D.3 · Table S11 (HTML A4.T11), data row 9 (NT-v2 [ 121 ]), column 6 (MRL (LBKWK))
Pipeline: NT-v2Protocol: MIMIC mRNABench probes MRL (Sugimoto): MRL (Sugimoto) prediction
Dataset subset: mRNABench MRL (Sugimoto) as reported in MIMIC Table S11 (MIMIC mRNABench probes split)
0.297 Pearson
unitless · higher

Uncertainty: Not yet extracted

Coverage: Not reported scored / Not reported eligible

Result quoted from another source · Source checked
Methods, coverage and source

NT-v2 on MIMIC mRNABench probes MRL (Sugimoto): MRL (Sugimoto) prediction

MIMIC uses concatenated register tokens and mean-pooled RNA-track representations with supervised probes. Nucleotide/codon/amino-acid conditioning is selected per task on validation data; predictions are averaged across selected modality subsets. Comparator scores are quoted from prior mRNABench work.

Aggregation: Not reported

MIMIC v1: Table S11 and Appendix D.3 · Table S11 (HTML A4.T11), data row 9 (NT-v2 [ 121 ]), column 7 (MRL (Sugimoto))
Pipeline: NT-v2Protocol: MIMIC mRNABench probes Protein Localization: Protein Localization prediction
Dataset subset: mRNABench Protein Localization as reported in MIMIC Table S11 (MIMIC mRNABench probes split)
0.336 AUPRC
fraction · higher

Uncertainty: Not yet extracted

Coverage: Not reported scored / Not reported eligible

Result quoted from another source · Source checked
Methods, coverage and source

NT-v2 on MIMIC mRNABench probes Protein Localization: Protein Localization prediction

MIMIC uses concatenated register tokens and mean-pooled RNA-track representations with supervised probes. Nucleotide/codon/amino-acid conditioning is selected per task on validation data; predictions are averaged across selected modality subsets. Comparator scores are quoted from prior mRNABench work.

Aggregation: Not reported

MIMIC v1: Table S11 and Appendix D.3 · Table S11 (HTML A4.T11), data row 9 (NT-v2 [ 121 ]), column 4 (Protein Localization)
Pipeline: NT-v2Protocol: MIMIC mRNABench probes RNA Half Life: RNA Half Life prediction
Dataset subset: mRNABench RNA Half Life as reported in MIMIC Table S11 (MIMIC mRNABench probes split)
0.53 Pearson
unitless · higher

Uncertainty: Not yet extracted

Coverage: Not reported scored / Not reported eligible

Result quoted from another source · Source checked
Methods, coverage and source

NT-v2 on MIMIC mRNABench probes RNA Half Life: RNA Half Life prediction

MIMIC uses concatenated register tokens and mean-pooled RNA-track representations with supervised probes. Nucleotide/codon/amino-acid conditioning is selected per task on validation data; predictions are averaged across selected modality subsets. Comparator scores are quoted from prior mRNABench work.

Aggregation: Not reported

MIMIC v1: Table S11 and Appendix D.3 · Table S11 (HTML A4.T11), data row 9 (NT-v2 [ 121 ]), column 8 (RNA Half Life)
Pipeline: NT-v2Protocol: MIMIC mRNABench probes RNA Localization: RNA Localization prediction
Dataset subset: mRNABench RNA Localization as reported in MIMIC Table S11 (MIMIC mRNABench probes split)
0.719 AUPRC
fraction · higher

Uncertainty: Not yet extracted

Coverage: Not reported scored / Not reported eligible

Result quoted from another source · Source checked
Methods, coverage and source

NT-v2 on MIMIC mRNABench probes RNA Localization: RNA Localization prediction

MIMIC uses concatenated register tokens and mean-pooled RNA-track representations with supervised probes. Nucleotide/codon/amino-acid conditioning is selected per task on validation data; predictions are averaged across selected modality subsets. Comparator scores are quoted from prior mRNABench work.

Aggregation: Not reported

MIMIC v1: Table S11 and Appendix D.3 · Table S11 (HTML A4.T11), data row 9 (NT-v2 [ 121 ]), column 5 (RNA Localization)

Source checking is not independent reproduction. Release 2026-11-21-c8bea25e5abb.

Use this pipeline

How it works, versions and access

Underlying model: Nucleotide Transformer. Results on this page belong to this pipeline and its evaluated settings.

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.

4 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-11-21-c8bea25e5abb
Property and statementOriginal source and locationReview and provenance
Description
Embeds mature mRNA with a Nucleotide Transformer model reported as NT-v2, from a family of transformer DNA foundation models (50 million to 2.5 billion parameters, integrating information from 3,202 human genomes and 850 genomes of other species), for the mRNABench property prediction tasks.
Individual claims
MIMIC v1: Table S11 and Appendix D.3

Original source ↗

coverage-source-mimic-2026-v1-html: Appendix D.3 and Table S11; expansion-p3-mrnabench-2025: abstract; qa-descriptions-20261010-source-nt-v2: abstract

Shared locator for this statement’s cited sources; not a separate locator for each citation.

Version: arXiv version 1
Retrieved: 2026-09-23

conflicting claim

["automated-source-review","ai-assisted-source-review"] · 2026-10-10

Audit details

Description written from the cited source text; provenance text moved to attributes.source_locator. Independent review 2026-10-10: corrected. Reworded so the parameter range and training genomes describe the family, not the evaluated model.

Conflicting source claim: "mRNABench evaluation of mature mRNA embeddings from a Nucleotide Transformer model reported as NT-v2, from a family of transformer DNA foundation models (50 million to 2.5 billion parameters, integrating information from 3,202 human genomes and 850 genomes of other species), with task scores quoted from prior work in MIMIC Table S11.". Claim record status: source_checked.

Field: description

Claim: metadata-correction-f089d399899322c2b49d

Source artifact SHA-256: 6408e6fae37567837ca53bd8a1ebef6cf7d3787bbf00dae7556e1b62822ffff0

Hash scope: Hash scope not separately documented; inspect source record

Inspected artifact

Description
Embeds mature mRNA with a Nucleotide Transformer model reported as NT-v2, from a family of transformer DNA foundation models (50 million to 2.5 billion parameters, integrating information from 3,202 human genomes and 850 genomes of other species), for the mRNABench property prediction tasks.
Individual claims
mRNABench: A curated benchmark for mature mRNA property and function prediction

Original source ↗

coverage-source-mimic-2026-v1-html: Appendix D.3 and Table S11; expansion-p3-mrnabench-2025: abstract; qa-descriptions-20261010-source-nt-v2: abstract

Shared locator for this statement’s cited sources; not a separate locator for each citation.

Version: preprint archived 2025-07-08
Retrieved: 2026-09-16T10:41:16.497221+00:00

conflicting claim

["automated-source-review","ai-assisted-source-review"] · 2026-10-10

Audit details

Description written from the cited source text; provenance text moved to attributes.source_locator. Independent review 2026-10-10: corrected. Reworded so the parameter range and training genomes describe the family, not the evaluated model.

Conflicting source claim: "mRNABench evaluation of mature mRNA embeddings from a Nucleotide Transformer model reported as NT-v2, from a family of transformer DNA foundation models (50 million to 2.5 billion parameters, integrating information from 3,202 human genomes and 850 genomes of other species), with task scores quoted from prior work in MIMIC Table S11.". Claim record status: source_checked.

Field: description

Claim: metadata-correction-f089d399899322c2b49d

Source artifact SHA-256: 79f6264ee883535203c63a313547e7c57baa85585f76b42f8d899eb17fb7e600

Hash scope: Exact retrieved primary paper artifact bytes.

Inspected artifact

Description
Embeds mature mRNA with a Nucleotide Transformer model reported as NT-v2, from a family of transformer DNA foundation models (50 million to 2.5 billion parameters, integrating information from 3,202 human genomes and 850 genomes of other species), for the mRNABench property prediction tasks.
Individual claims
Abstract: Nucleotide Transformer: building and evaluating robust foundation models for human genomics

Original source ↗

coverage-source-mimic-2026-v1-html: Appendix D.3 and Table S11; expansion-p3-mrnabench-2025: abstract; qa-descriptions-20261010-source-nt-v2: abstract

Shared locator for this statement’s cited sources; not a separate locator for each citation.

Version: Europe PMC core record MED/39609566 (title and abstract), Nature methods 2025
Retrieved: 2026-10-10T06:44:13Z

conflicting claim

["automated-source-review","ai-assisted-source-review"] · 2026-10-10

Audit details

Description written from the cited source text; provenance text moved to attributes.source_locator. Independent review 2026-10-10: corrected. Reworded so the parameter range and training genomes describe the family, not the evaluated model.

Conflicting source claim: "mRNABench evaluation of mature mRNA embeddings from a Nucleotide Transformer model reported as NT-v2, from a family of transformer DNA foundation models (50 million to 2.5 billion parameters, integrating information from 3,202 human genomes and 850 genomes of other species), with task scores quoted from prior work in MIMIC Table S11.". Claim record status: source_checked.

Field: description

Claim: metadata-correction-f089d399899322c2b49d

Source artifact SHA-256: 94fd1bf1c8ac29a4d037b6f9bdc7eda76a47f1d1efa67082b7074f7a219e817a

Hash scope: SHA-256 of one retrieval of a Europe PMC REST search response. The response is not byte-stable: on 2026-10-10 a second retrieval of 10 of these 31 records differed only in JSON key order and parsed to identical content. The title and abstractText fields are the evidence.

Inspected artifact

Relationship: uses model
discovery-model-nucleotide-transformer
Individual claims
MIMIC v1: Table S11 and Appendix D.3

Original source ↗

Table S11, row NT-v2 [ 121 ]; Appendix D.3

Version: arXiv version 1
Retrieved: 2026-09-23

source checked

["automated-source-review"] · 2026-09-23

Audit details

Source-backed evaluated identity only; no independent reproduction.

Field: links:uses_model:discovery-model-nucleotide-transformer

Claim: mimic-2026-mrnabench-method-nt-v2-discovery-model-nucleotide-transformer-identity-claim

Source artifact SHA-256: 6408e6fae37567837ca53bd8a1ebef6cf7d3787bbf00dae7556e1b62822ffff0

Hash scope: Hash scope not separately documented; inspect source record

Inspected artifact

Sources and history

Release 2026-11-21-c8bea25e5abb · Record review: source checked

3 source records and release historyDownload this release (gzip)
Technical metadata and record identifiers

Technical metadata for this record is tracked under stable ID mimic-2026-mrnabench-method-nt-v2.

Source locator: Table S11, row NT-v2 [ 121 ]; Appendix D.3; Prior mRNABench evaluated pipeline, quoted in MIMIC Table S11; not a new execution

Related records

How to cite

NT-v2. Rewire biological model benchmark database, release 2026-11-21-c8bea25e5abb. https://benchmarks.rewirebio.io/database/pipeline/mimic-2026-mrnabench-method-nt-v2/

Release 2026-11-21-c8bea25e5abb. Rankings and values change between releases, so keep the release id in the citation. Cite the original publication as well for any value you reuse: each result links its source.

The compilation and annotations made by Rewire in this release are offered under CC BY 4.0; values and text taken from cited sources stay under the terms of those sources and are not covered by this licence. Data licence

Suggest a correction