rewire.it
Task

DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

Cell-type-specific element classification, IMR90. Scored with AUROC on ENCODE chromatin accessibility peaks in five cell lines. One-against-rest AUROC for IMR90 accessible elements.

14 evaluations · 14 metric rows

Overview

Cell-type-specific element classification, IMR90. Scored with AUROC on ENCODE chromatin accessibility peaks in five cell lines. One-against-rest AUROC for IMR90 accessible elements.

Consult the linked sources for architecture or protocol details. Missing evidence is not evidence of a missing capability.

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.

Explore all linked results

Run instructions

No runnable recipe has been reviewed for this task. Dataset access, model requirements, licences and compute requirements must be checked against its sources before execution.

A task describes a biological question. Choose a linked protocol to obtain concrete split and scoring instructions.

Published comparisons

Explore the results reported under one evaluation protocol. Each figure keeps its source, dataset and metric together; it is not a ranking across studies. The pooled view gathers every source table that reports the same metric and names what it does not hold constant.

DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

auroc (fraction) · Higher values are better for this metric.

Every method DART-Eval reports on Cell-type-specific element classification, IMR90, scored with AUROC on ENCODE chromatin accessibility peaks in five cell lines.

Evaluation protocol · ENCODE chromatin accessibility peaks in five cell lines (DART-Eval split)

  1. Caduceus (probed) · Configuration · Author-reported evaluation0.576
  2. DNABERT-2 (probed) · Configuration · Author-reported evaluation0.691
  3. GENA-LM (probed) · Configuration · Author-reported evaluation0.714
  4. HyenaDNA (probed) · Configuration · Author-reported evaluation0.882
  5. Mistral-DNA (probed) · Configuration · Author-reported evaluation0.643
  6. Caduceus (fine-tuned) · Configuration · Author-reported evaluation0.929
  7. DNABERT-2 (fine-tuned) · Configuration · Author-reported evaluation0.922
  8. GENA-LM (fine-tuned) · Configuration · Author-reported evaluation0.911
  9. HyenaDNA (fine-tuned) · Configuration · Author-reported evaluation0.908
  10. Probing-head-like (ab initio) · Method · Author-reported evaluation0.807
  11. ChromBPNet-like (ab initio) · Method · Author-reported evaluation0.921

Source order is preserved. Plotted marks show point estimates; uncertainty, where reported, is retained in the printed values and table. Differences do not establish statistical significance.

DART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, column(IMR90 AUROC)
Values, uncertainty and evidence
auroc: original source values
Tested entityPrinted valueUncertaintyEvidence
Caduceus (probed) · Configuration0.576 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed Caduceus), column(CTS-IMR90)
DNABERT-2 (probed) · Configuration0.691 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed DNABERT-2), column(CTS-IMR90)
GENA-LM (probed) · Configuration0.714 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed GENA-LM), column(CTS-IMR90)
HyenaDNA (probed) · Configuration0.882 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed HyenaDNA), column(CTS-IMR90)
Mistral-DNA (probed) · Configuration0.643 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed Mistral-DNA), column(CTS-IMR90)
Nucleotide Transformer (probed) · Configuration0.779 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed Nucleotide Transformer), column(CTS-IMR90)
Caduceus (fine-tuned) · Configuration0.929 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned Caduceus), column(CTS-IMR90)
DNABERT-2 (fine-tuned) · Configuration0.922 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned DNABERT-2), column(CTS-IMR90)
GENA-LM (fine-tuned) · Configuration0.911 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned GENA-LM), column(CTS-IMR90)
HyenaDNA (fine-tuned) · Configuration0.908 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned HyenaDNA), column(CTS-IMR90)
Mistral-DNA (fine-tuned) · Configuration0.748 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned Mistral-DNA), column(CTS-IMR90)
Nucleotide Transformer (fine-tuned) · Configuration0.920 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned Nucleotide Transformer), column(CTS-IMR90)
Probing-head-like (ab initio) · Method0.807 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Ab initio Probing-head-like), column(CTS-IMR90)
ChromBPNet-like (ab initio) · Method0.921 fractionNot reportedAuthor-reported evaluation · source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Ab initio ChromBPNet-like), column(CTS-IMR90)
Scope and limitations
  • Author-reported numbers, source checked but not independently reproduced.
  • The evaluation setting is part of the method name: a zero-shot, probed and fine-tuned run of the same model are different entries.
  • Metrics and datasets differ between tasks, so these figures cannot be averaged into one score.

Source transcription and grouping reviewed by automated source review on 2026-09-18. These experiments were not independently reproduced by rewire.

Tested entities and results

Release 2026-09-17-134cd1815de8 · 14 evaluations · 14 metric rows. Different protocols are not a single leaderboard. Where several source tables report the same metric, the published comparisons above offer a pooled view that names what it does not hold constant.

Results grouped by the exact reported evaluation
Metric and findingCoverage and uncertaintyEvidence
Caduceus (fine-tuned) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.929 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned Caduceus), column(CTS-IMR90)

Source checking is not independent reproduction.

Caduceus (probed) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.576 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed Caduceus), column(CTS-IMR90)

Source checking is not independent reproduction.

ChromBPNet-like (ab initio) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.921 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Ab initio ChromBPNet-like), column(CTS-IMR90)

Source checking is not independent reproduction.

DNABERT-2 (fine-tuned) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.922 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned DNABERT-2), column(CTS-IMR90)

Source checking is not independent reproduction.

DNABERT-2 (probed) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.691 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed DNABERT-2), column(CTS-IMR90)

Source checking is not independent reproduction.

GENA-LM (fine-tuned) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.911 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned GENA-LM), column(CTS-IMR90)

Source checking is not independent reproduction.

GENA-LM (probed) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.714 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed GENA-LM), column(CTS-IMR90)

Source checking is not independent reproduction.

HyenaDNA (fine-tuned) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.908 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned HyenaDNA), column(CTS-IMR90)

Source checking is not independent reproduction.

HyenaDNA (probed) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.882 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed HyenaDNA), column(CTS-IMR90)

Source checking is not independent reproduction.

Mistral-DNA (fine-tuned) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.748 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned Mistral-DNA), column(CTS-IMR90)

Source checking is not independent reproduction.

Mistral-DNA (probed) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.643 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed Mistral-DNA), column(CTS-IMR90)

Source checking is not independent reproduction.

Nucleotide Transformer (fine-tuned) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.920 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Fine-Tuned Nucleotide Transformer), column(CTS-IMR90)

Source checking is not independent reproduction.

Nucleotide Transformer (probed) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.779 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Probed Nucleotide Transformer), column(CTS-IMR90)

Source checking is not independent reproduction.

Probing-head-like (ab initio) on DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90

One-against-rest AUROC for IMR90 accessible elements.

Author-reported evaluation · Evaluation metadata: source checked

0.807 auroc

Unit: fraction · Direction: higher

Uncertainty: Not reported

Scored: Not reported · Eligible: Not reported

source checkedDART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA · Table 4, row(Ab initio Probing-head-like), column(CTS-IMR90)

Source checking is not 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

Claims, original sources and review scope · Release 2026-09-17-134cd1815de8
Property and statementOriginal source and locationReview and provenance
Relationship: part of

discovery-benchmark-dart-eval

Individual claims
DART-Eval: A Comprehensive DNA Language Model Evaluation Benchmark on Regulatory DNA

Original source ↗

Table 4, column(IMR90 AUROC)

Version: 2412.05430v1
Retrieved: 2026-09-17T07:56:09.182117+00:00

source checked

automated source review · 2026-09-18

Audit details

Primary-source transcription with no human sign-off and no independent reproduction.

Field: links:part_of:discovery-benchmark-dart-eval

Claim: dart-eval-association-cts-imr90

Source artifact SHA-256: 4194b137ba55c9a2c269d119a9afec6ae1bb0feaf17d91433ae483c41221a56b

Hash scope: Exact retrieved primary paper artifact bytes.

Inspected artifact

Sources and history

Release 2026-09-17-134cd1815de8 · Record review: source checked

1 source records and release historyDownload this release
Technical metadata and extraction receipts

Stable ID: dart-eval-task-cts-imr90

areas
dna-genomes
tasks
Cell-type-specific element classification, IMR90
metric
AUROC
metric direction
higher
dataset
ENCODE chromatin accessibility peaks in five cell lines
protocol
One-against-rest AUROC for IMR90 accessible elements.
source locator
Table 4, column(IMR90 AUROC)
comparison panels
id: dart-eval-panel-cts-imr90; title: DART-Eval CTS-IMR90: Cell-type-specific element classification, IMR90; protocol id: dart-eval-task-cts-imr90; dataset id: dart-eval-dataset-encode-chromatin-accessibility-peaks-in-five-cell-lines; metric: auroc; unit: fraction; direction: higher; result ids: dart-eval-result-caduceus-probed-cts-imr90-auroc; dart-eval-result-dnabert-2-probed-cts-imr90-auroc; dart-eval-result-gena-lm-probed-cts-imr90-auroc; dart-eval-result-hyenadna-probed-cts-imr90-auroc; dart-eval-result-mistral-dna-probed-cts-imr90-auroc; dart-eval-result-nucleotide-transformer-probed-cts-imr90-auroc; dart-eval-result-caduceus-fine-tuned-cts-imr90-auroc; dart-eval-result-dnabert-2-fine-tuned-cts-imr90-auroc; dart-eval-result-gena-lm-fine-tuned-cts-imr90-auroc; dart-eval-result-hyenadna-fine-tuned-cts-imr90-auroc; dart-eval-result-mistral-dna-fine-tuned-cts-imr90-auroc; dart-eval-result-nucleotide-transformer-fine-tuned-cts-imr90-auroc; dart-eval-result-probing-head-like-ab-initio-cts-imr90-auroc; dart-eval-result-chrombpnet-like-ab-initio-cts-imr90-auroc; source ids: evidence-expansion-p2-evidence-discovery-final-dart-4194b137ba55; source locator: Table 4, column(IMR90 AUROC); context: Every method DART-Eval reports on Cell-type-specific element classification, IMR90, scored with AUROC on ENCODE chromatin accessibility peaks in five cell lines.; caveats: Author-reported numbers, source checked but not independently reproduced.; The evaluation setting is part of the method name: a zero-shot, probed and fine-tuned run of the same model are different entries.; Metrics and datasets differ between tasks, so these figures cannot be averaged into one score.; review: method: automated_source_review; date: 2026-09-18
Related records

Suggest a correction