NT benchmark H3K4me1, fine-tuning (PEFT); Nature Methods 2025

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.

14 recorded evaluations, 14 metric rows. A comparison chart has not yet been validated for these results. The table retains the individual findings and their sources.

View coverage and remaining gaps across all benchmarks

Results

Results are available, but no reviewed comparison panel is linked in this release.

All evaluations

14 evaluations · 14 results. Different protocols are not a single leaderboard.

Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) · Dataset: NT benchmark H3K4me1

Sorted by MCC (median over 10 cross-validation folds) (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.

Tested configuration	MCC (median over 10 cross-validation folds)
Enformer, parameter-efficient fine-tuning (NT benchmark)Evaluation	0.504 (± 0.021) (best in column)
NT-Multispecies-v2 (500M), parameter-efficient fine-tuning (NT benchmark)Evaluation	0.497 (± 0.009)
NT-1000G (2.5B), parameter-efficient fine-tuning (NT benchmark)Evaluation	0.491 (± 0.015)
NT-Multispecies-v2 (250M), parameter-efficient fine-tuning (NT benchmark)Evaluation	0.491 (± 0.006)
DNABERT-2, parameter-efficient fine-tuning (NT benchmark)Evaluation	0.49 (± 0.008)
NT-Multispecies-v2 (100M), parameter-efficient fine-tuning (NT benchmark)Evaluation	0.485 (± 0.011)
NT-Multispecies (2.5B), parameter-efficient fine-tuning (NT benchmark)Evaluation	0.481 (± 0.012)
NT-Multispecies-v2 (50M), parameter-efficient fine-tuning (NT benchmark)Evaluation	0.478 (± 0.012)
NT-HumanRef (500M), parameter-efficient fine-tuning (NT benchmark)Evaluation	0.468 (± 0.007)
NT-1000G (500M), parameter-efficient fine-tuning (NT benchmark)Evaluation	0.466 (± 0.006)
HyenaDNA-1KB, parameter-efficient fine-tuning (NT benchmark)Evaluation	0.43 (± 0.014)
HyenaDNA-32KB, parameter-efficient fine-tuning (NT benchmark)Evaluation	0.423 (± 0.016)
NT 2.5B Randomly Initialized, trained from scratch (NT benchmark)Evaluation	0.271 (± 0.02)
NT 500M Randomly Initialized, trained from scratch (NT benchmark)Evaluation	0.228 (± 0.018)

All 14 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: DNABERT-2, parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.49 (± 0.008) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.008 Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source DNABERT-2 on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'DNABERT-2', column 'H3K4me1'
Configuration: Enformer, parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.504 (± 0.021) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.021 Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source Enformer on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'Enformer', column 'H3K4me1'
Configuration: HyenaDNA-1KB, parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.43 (± 0.014) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.014 Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source HyenaDNA-1KB on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'HyenaDNA-1KB', column 'H3K4me1'
Configuration: HyenaDNA-32KB, parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.423 (± 0.016) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.016 Coverage: Not reported scored / Not reported eligible	Independent external evaluation · Source checked Methods, coverage and source HyenaDNA-32KB on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'HyenaDNA-32KB', column 'H3K4me1'
Configuration: NT-1000G (2.5B), parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.491 (± 0.015) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.015 Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source NT-1000G (2.5B) on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'NT-1000G (2.5B)', column 'H3K4me1'
Configuration: NT-1000G (500M), parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.466 (± 0.006) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.006 Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source NT-1000G (500M) on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'NT-1000G (500M)', column 'H3K4me1'
Configuration: NT 2.5B Randomly Initialized, trained from scratch (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.271 (± 0.02) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.02 Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source NT 2.5B Randomly Initialized on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'NT 2.5B Randomly Initialized', column 'H3K4me1'
Configuration: NT 500M Randomly Initialized, trained from scratch (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.228 (± 0.018) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.018 Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source NT 500M Randomly Initialized on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'NT 500M Randomly Initialized', column 'H3K4me1'
Configuration: NT-HumanRef (500M), parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.468 (± 0.007) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.007 Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source NT-HumanRef (500M) on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'NT-HumanRef (500M)', column 'H3K4me1'
Configuration: NT-Multispecies (2.5B), parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.481 (± 0.012) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.012 Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source NT-Multispecies (2.5B) on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'NT-Multispecies (2.5B)', column 'H3K4me1'
Configuration: NT-Multispecies-v2 (100M), parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.485 (± 0.011) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.011 Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source NT-Multispecies-v2 (100M) on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'NT-Multispecies-v2 (100M)', column 'H3K4me1'
Configuration: NT-Multispecies-v2 (250M), parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.491 (± 0.006) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.006 Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source NT-Multispecies-v2 (250M) on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'NT-Multispecies-v2 (250M)', column 'H3K4me1'
Configuration: NT-Multispecies-v2 (500M), parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.497 (± 0.009) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.009 Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source NT-Multispecies-v2 (500M) on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'NT-Multispecies-v2 (500M)', column 'H3K4me1'
Configuration: NT-Multispecies-v2 (50M), parameter-efficient fine-tuning (NT benchmark)	Protocol: NT benchmark H3K4me1, fine-tuning (PEFT) Dataset: NT benchmark H3K4me1	0.478 (± 0.012) MCC (median over 10 cross-validation folds) unitless · higher Uncertainty: SD 0.012 Coverage: Not reported scored / Not reported eligible	Author-reported evaluation · Source checked Methods, coverage and source NT-Multispecies-v2 (50M) on NT benchmark H3K4me1 (fine-tuning) lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning Aggregation: Not reported Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Supplementary Table 7, row 'NT-Multispecies-v2 (50M)', column 'H3K4me1'

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

Methods and evaluation design

Procedure, tasks and evaluated configurations

Recorded evaluations

Each evaluation records what was tested and under which conditions.

Explore all linked results

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: 10
External evaluations: 4

Literature evidence is not a Rewire measurement. Executed but unpublished runs and private review status are not included.

Null control

Proposed control: requires review

Select a task-valid null control after reviewing inputs and metric

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

Select an upstream conventional reference after reviewing the full protocol

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 (gzip) · Model evaluation matrix (gzip) · Source table (gzip) · Release and checksums (gzip)

Coverage is derived from release 2026-10-22-1b7cf027e4f6. 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.

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

Nucleotide Transformer: Supplementary Information (Supplementary Tables 1-12) · Original source · Nature Methods 22, 287-297 (published online 2024-11-28); Supplementary Information PDF (MOESM1, 25 pages)

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: lit-genomics-20261011-ntbench-protocol-h3k4me1-fine-tuning

areas: dna-genomes
contexts: research
protocol: fine-tuning (PEFT); MCC per fold, reported as the median of 10 cross-validation folds with standard deviation. BPNet models are fully trained CNNs and the Raw Probe uses probing in both tables.
version: Nature Methods 2025
metric: matthews-correlation-coefficient
metric direction: higher
unit: unitless
limitations: Developer benchmark: NT authors ran every model.; Cross-validation median; the dataset 'test' column of Supplementary Table 5 is not separately reported.
source locator: Supplementary Table 6 (probing) or 7 (fine-tuning) caption; Supplementary Table 5

Related records

part of: Nucleotide Transformer benchmark (18 tasks)
uses data: NT benchmark H3K4me1
assessment: DNABERT-2 on NT benchmark H3K4me1 (fine-tuning)
assessment: Enformer on NT benchmark H3K4me1 (fine-tuning)
assessment: HyenaDNA-1KB on NT benchmark H3K4me1 (fine-tuning)
assessment: HyenaDNA-32KB on NT benchmark H3K4me1 (fine-tuning)
assessment: NT-1000G (2.5B) on NT benchmark H3K4me1 (fine-tuning)
assessment: NT-1000G (500M) on NT benchmark H3K4me1 (fine-tuning)
assessment: NT 2.5B Randomly Initialized on NT benchmark H3K4me1 (fine-tuning)
assessment: NT 500M Randomly Initialized on NT benchmark H3K4me1 (fine-tuning)
assessment: NT-HumanRef (500M) on NT benchmark H3K4me1 (fine-tuning)
assessment: NT-Multispecies (2.5B) on NT benchmark H3K4me1 (fine-tuning)
assessment: NT-Multispecies-v2 (100M) on NT benchmark H3K4me1 (fine-tuning)
assessment: NT-Multispecies-v2 (250M) on NT benchmark H3K4me1 (fine-tuning)
assessment: NT-Multispecies-v2 (500M) on NT benchmark H3K4me1 (fine-tuning)
assessment: NT-Multispecies-v2 (50M) on NT benchmark H3K4me1 (fine-tuning)