Source

Source Data Fig. 2, Ahlmann-Eltze et al. 2025

Evidence concern: excluded from comparisons

The Fig. 2 legend gives "134, 210 and 24 unseen single perturbations across two test-training splits". In Panel A these are the split 1 counts only (Replogle K562 134, Replogle RPE1 210, Adamson 24 rows per method); split 2 has 128, 203 and 24, so the two-split totals are 262, 413 and 48. The Fig. 2a means recorded here use all Panel A rows; whether the plotted red lines use both splits is not stated beyond the sheet itself. Reviewer check 2026-10-11: using split 1 rows only would change the per-model means by up to 0.61 (L2 distance) and 0.06 (Pearson delta), and would change the model with the lowest mean L2 distance on Adamson and Replogle RPE1, so the choice is material.Source Data Fig. 2, Ahlmann-Eltze et al. 2025 · Fig. 2 legend versus Source Data Fig. 2, sheet "Panel A"

Methods restricts the single-perturbation analysis to perturbations predicted by all models: 73 for Adamson, 398 for Replogle K562 and 629 for Replogle RPE1. No sheet of Source Data Fig. 2 has these counts (Panel A: 48, 262 and 413 per method over two splits; Panel C: 20, 214 and 224 test or val rows per method).Source Data Fig. 2, Ahlmann-Eltze et al. 2025 · Methods 'Single perturbation benchmark setup' paragraph 4 versus Source Data Fig. 2, sheets "Panel A" and "Panel C"

The test and val rows in Panel C (20, 214 and 224 per method for Adamson, Replogle K562 and Replogle RPE1) match neither the Fig. 2 legend counts nor the Methods counts. Fig. 2c was not extracted.Source Data Fig. 2, Ahlmann-Eltze et al. 2025 · Source Data Fig. 2, sheet "Panel C" versus the Fig. 2 legend and Methods

The main text says CPA was not included in the single-perturbation benchmark, but method "cpa" has rows in Panel C. Fig. 2c was not extracted.Source Data Fig. 2, Ahlmann-Eltze et al. 2025 · Main text paragraph 16 versus Source Data Fig. 2, sheet "Panel C"

Nature Methods · 2025 · Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article · Peer-reviewed

DOI: 10.1038/s41592-025-02772-6 Original source

1837063 bytes. Each row of sheet "Panel A" is one held-out perturbation in one test-training split (column seed), labelled test or val in column train.

What this source contributed

Records in this release that cite this source. A record can cite several sources.

21 evaluations
42 results
3 evidence claims

Results

Values as the source printed them. Compare results only under compatible protocols, datasets and metrics.

Value	Metric	Model or configuration	Dataset	Evaluation
3.867 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	GEARS (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	GEARS on Adamson (Ahlmann-Eltze et al. 2025)
0.726 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	GEARS (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	GEARS on Adamson (Ahlmann-Eltze et al. 2025)
4.079 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	GEARS (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	GEARS on Replogle K562 (Ahlmann-Eltze et al. 2025)
0.388 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	GEARS (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	GEARS on Replogle K562 (Ahlmann-Eltze et al. 2025)
6.113 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	GEARS (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	GEARS on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
0.636 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	GEARS (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	GEARS on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
3.908 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Geneformer (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	Geneformer on Adamson (Ahlmann-Eltze et al. 2025)
0.726 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Geneformer (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	Geneformer on Adamson (Ahlmann-Eltze et al. 2025)
4.236 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Geneformer (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Geneformer on Replogle K562 (Ahlmann-Eltze et al. 2025)
0.404 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Geneformer (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Geneformer on Replogle K562 (Ahlmann-Eltze et al. 2025)
5.494 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Geneformer (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Geneformer on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
0.678 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Geneformer (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Geneformer on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
4.081 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Linear model (G and P from training data) (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	Linear model (G and P from training data) on Adamson (Ahlmann-Eltze et al. 2025)
0.717 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Linear model (G and P from training data) (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	Linear model (G and P from training data) on Adamson (Ahlmann-Eltze et al. 2025)
4.097 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Linear model (G and P from training data) (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Linear model (G and P from training data) on Replogle K562 (Ahlmann-Eltze et al. 2025)
0.436 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Linear model (G and P from training data) (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Linear model (G and P from training data) on Replogle K562 (Ahlmann-Eltze et al. 2025)
5.390 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Linear model (G and P from training data) (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Linear model (G and P from training data) on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
0.680 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Linear model (G and P from training data) (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Linear model (G and P from training data) on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
3.936 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Mean (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	Mean on Adamson (Ahlmann-Eltze et al. 2025)
0.724 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Mean (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	Mean on Adamson (Ahlmann-Eltze et al. 2025)
4.193 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Mean (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Mean on Replogle K562 (Ahlmann-Eltze et al. 2025)
0.408 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Mean (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Mean on Replogle K562 (Ahlmann-Eltze et al. 2025)
5.534 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Mean (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Mean on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
0.678 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	Mean (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	Mean on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
3.878 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scBERT (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	scBERT on Adamson (Ahlmann-Eltze et al. 2025)
0.728 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scBERT (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	scBERT on Adamson (Ahlmann-Eltze et al. 2025)
4.192 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scBERT (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	scBERT on Replogle K562 (Ahlmann-Eltze et al. 2025)
0.408 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scBERT (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	scBERT on Replogle K562 (Ahlmann-Eltze et al. 2025)
5.528 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scBERT (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	scBERT on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
0.678 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scBERT (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	scBERT on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
4.951 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scGPT (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	scGPT on Adamson (Ahlmann-Eltze et al. 2025)
0.595 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scGPT (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	scGPT on Adamson (Ahlmann-Eltze et al. 2025)
4.437 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scGPT (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	scGPT on Replogle K562 (Ahlmann-Eltze et al. 2025)
0.316 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scGPT (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	scGPT on Replogle K562 (Ahlmann-Eltze et al. 2025)
5.991 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scGPT (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	scGPT on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
0.588 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	scGPT (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	scGPT on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
3.936 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	UCE (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	UCE on Adamson (Ahlmann-Eltze et al. 2025)
0.724 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	UCE (Ahlmann-Eltze et al. 2025)	Adamson et al. 2016 Perturb-seq, K562 CRISPRi single perturbations (GEARS distribution, Ahlmann-Eltze et al. 2025)	UCE on Adamson (Ahlmann-Eltze et al. 2025)
4.200 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	UCE (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	UCE on Replogle K562 (Ahlmann-Eltze et al. 2025)
0.409 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	UCE (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, K562 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	UCE on Replogle K562 (Ahlmann-Eltze et al. 2025)
5.507 unit-unreported	L2 distance (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	UCE (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	UCE on Replogle RPE1 (Ahlmann-Eltze et al. 2025)
0.680 unitless	Pearson delta (1,000 most highly expressed control genes; mean over held-out perturbations and splits)	UCE (Ahlmann-Eltze et al. 2025)	Replogle et al. 2022 Perturb-seq essential-gene screen, RPE1 CRISPRi (GEARS distribution, Ahlmann-Eltze et al. 2025)	UCE on Replogle RPE1 (Ahlmann-Eltze et al. 2025)

Claims

Statements this source supports, each with where in the source it was checked.

Relevance of Adamson: expression prediction for unseen single perturbations (Ahlmann-Eltze et al. 2025 Fig. 2a and Extended Data Fig. 8a) to "Assess models for genetic perturbation experiments": Assessed by Adamson: expression prediction for unseen single perturbations (Ahlmann-Eltze et al. 2025 Fig. 2a and Extended Data Fig. 8a) (perturbation-ahlmann-20261011-source-article: Fig. 2 and Extended Data Fig. 8 legends and Methods; perturbation-ahlmann-20261011-source-data-fig2: sheet "Panel A"; perturbation-ahlmann-20261011-source-data-ed-fig8: sheet "Panel A")
Relevance of Replogle K562: expression prediction for unseen single perturbations (Ahlmann-Eltze et al. 2025 Fig. 2a and Extended Data Fig. 8a) to "Assess models for genetic perturbation experiments": Assessed by Replogle K562: expression prediction for unseen single perturbations (Ahlmann-Eltze et al. 2025 Fig. 2a and Extended Data Fig. 8a) (perturbation-ahlmann-20261011-source-article: Fig. 2 and Extended Data Fig. 8 legends and Methods; perturbation-ahlmann-20261011-source-data-fig2: sheet "Panel A"; perturbation-ahlmann-20261011-source-data-ed-fig8: sheet "Panel A")
Relevance of Replogle RPE1: expression prediction for unseen single perturbations (Ahlmann-Eltze et al. 2025 Fig. 2a and Extended Data Fig. 8a) to "Assess models for genetic perturbation experiments": Assessed by Replogle RPE1: expression prediction for unseen single perturbations (Ahlmann-Eltze et al. 2025 Fig. 2a and Extended Data Fig. 8a) (perturbation-ahlmann-20261011-source-article: Fig. 2 and Extended Data Fig. 8 legends and Methods; perturbation-ahlmann-20261011-source-data-fig2: sheet "Panel A"; perturbation-ahlmann-20261011-source-data-ed-fig8: sheet "Panel A")

Models, benchmarks and other records

Evaluations (21)

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.

21 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
attributes.artifact_sha256 2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35 Source metadata	Source Data Fig. 2, Ahlmann-Eltze et al. 2025 Original source ↗ No field-specific location recorded Version: Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article Retrieved: 2026-10-11T00:48:13Z	catalogued No individual claim review recorded Source has a recorded evidence concern. Consult its source page before using the claim. Audit details Field: `attributes.artifact_sha256` Source artifact SHA-256: `2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35` Hash scope: Hash scope not separately documented; inspect source record
Doi 10.1038/s41592-025-02772-6 Source metadata	Source Data Fig. 2, Ahlmann-Eltze et al. 2025 Original source ↗ No field-specific location recorded Version: Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article Retrieved: 2026-10-11T00:48:13Z	catalogued No individual claim review recorded Source has a recorded evidence concern. Consult its source page before using the claim. Audit details Field: `attributes.doi` Source artifact SHA-256: `2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35` Hash scope: Hash scope not separately documented; inspect source record
Evidence concerns 4 values source id perturbation-ahlmann-20261011-source-data-fig2 message The Fig. 2 legend gives "134, 210 and 24 unseen single perturbations across two test-training splits". In Panel A these are the split 1 counts only (Replogle K562 134, Replogle RPE1 210, Adamson 24 rows per method); split 2 has 128, 203 and 24, so the two-split totals are 262, 413 and 48. The Fig. 2a means recorded here use all Panel A rows; whether the plotted red lines use both splits is not stated beyond the sheet itself. Reviewer check 2026-10-11: using split 1 rows only would change the per-model means by up to 0.61 (L2 distance) and 0.06 (Pearson delta), and would change the model with the lowest mean L2 distance on Adamson and Replogle RPE1, so the choice is material. source locator Fig. 2 legend versus Source Data Fig. 2, sheet "Panel A" artifact sha256 2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35 reviewed at 2026-10-11T00:00:00Z review method ai-assisted-source-review source id perturbation-ahlmann-20261011-source-data-fig2 message Methods restricts the single-perturbation analysis to perturbations predicted by all models: 73 for Adamson, 398 for Replogle K562 and 629 for Replogle RPE1. No sheet of Source Data Fig. 2 has these counts (Panel A: 48, 262 and 413 per method over two splits; Panel C: 20, 214 and 224 test or val rows per method). source locator Methods 'Single perturbation benchmark setup' paragraph 4 versus Source Data Fig. 2, sheets "Panel A" and "Panel C" artifact sha256 2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35 reviewed at 2026-10-11T00:00:00Z review method ai-assisted-source-review source id perturbation-ahlmann-20261011-source-data-fig2 message The test and val rows in Panel C (20, 214 and 224 per method for Adamson, Replogle K562 and Replogle RPE1) match neither the Fig. 2 legend counts nor the Methods counts. Fig. 2c was not extracted. source locator Source Data Fig. 2, sheet "Panel C" versus the Fig. 2 legend and Methods artifact sha256 2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35 reviewed at 2026-10-11T00:00:00Z review method ai-assisted-source-review source id perturbation-ahlmann-20261011-source-data-fig2 message The main text says CPA was not included in the single-perturbation benchmark, but method "cpa" has rows in Panel C. Fig. 2c was not extracted. source locator Main text paragraph 16 versus Source Data Fig. 2, sheet "Panel C" artifact sha256 2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35 reviewed at 2026-10-11T00:00:00Z review method ai-assisted-source-review Source metadata	Source Data Fig. 2, Ahlmann-Eltze et al. 2025 Original source ↗ No field-specific location recorded Version: Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article Retrieved: 2026-10-11T00:48:13Z	catalogued No individual claim review recorded Source has a recorded evidence concern. Consult its source page before using the claim. Audit details Field: `attributes.evidence_concerns` Source artifact SHA-256: `2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35` Hash scope: Hash scope not separately documented; inspect source record
Licence CC-BY-4.0 Source metadata	Source Data Fig. 2, Ahlmann-Eltze et al. 2025 Original source ↗ No field-specific location recorded Version: Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article Retrieved: 2026-10-11T00:48:13Z	catalogued No individual claim review recorded Source has a recorded evidence concern. Consult its source page before using the claim. Audit details Field: `attributes.licence` Source artifact SHA-256: `2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35` Hash scope: Hash scope not separately documented; inspect source record
Media type application/vnd.openxmlformats-officedocument.spreadsheetml.sheet Source metadata	Source Data Fig. 2, Ahlmann-Eltze et al. 2025 Original source ↗ No field-specific location recorded Version: Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article Retrieved: 2026-10-11T00:48:13Z	catalogued No individual claim review recorded Source has a recorded evidence concern. Consult its source page before using the claim. Audit details Field: `attributes.media_type` Source artifact SHA-256: `2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35` Hash scope: Hash scope not separately documented; inspect source record
Publication status peer_reviewed Source metadata	Source Data Fig. 2, Ahlmann-Eltze et al. 2025 Original source ↗ No field-specific location recorded Version: Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article Retrieved: 2026-10-11T00:48:13Z	catalogued No individual claim review recorded Source has a recorded evidence concern. Consult its source page before using the claim. Audit details Field: `attributes.publication_status` Source artifact SHA-256: `2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35` Hash scope: Hash scope not separately documented; inspect source record
Retrieved at 2026-10-11T00:48:13Z Source metadata	Source Data Fig. 2, Ahlmann-Eltze et al. 2025 Original source ↗ No field-specific location recorded Version: Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article Retrieved: 2026-10-11T00:48:13Z	catalogued No individual claim review recorded Source has a recorded evidence concern. Consult its source page before using the claim. Audit details Field: `attributes.retrieved_at` Source artifact SHA-256: `2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35` Hash scope: Hash scope not separately documented; inspect source record
Scope note 1837063 bytes. Each row of sheet "Panel A" is one held-out perturbation in one test-training split (column seed), labelled test or val in column train. Source metadata	Source Data Fig. 2, Ahlmann-Eltze et al. 2025 Original source ↗ No field-specific location recorded Version: Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article Retrieved: 2026-10-11T00:48:13Z	catalogued No individual claim review recorded Source has a recorded evidence concern. Consult its source page before using the claim. Audit details Field: `attributes.scope_note` Source artifact SHA-256: `2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35` Hash scope: Hash scope not separately documented; inspect source record
Source location Article 'Data availability' and the 'Source Data Fig. 2' link under the figure; the article licence covers its supplementary material Source metadata	Source Data Fig. 2, Ahlmann-Eltze et al. 2025 Original source ↗ No field-specific location recorded Version: Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article Retrieved: 2026-10-11T00:48:13Z	catalogued No individual claim review recorded Source has a recorded evidence concern. Consult its source page before using the claim. Audit details Field: `attributes.source_locator` Source artifact SHA-256: `2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35` Hash scope: Hash scope not separately documented; inspect source record
Url https://static-content.springer.com/esm/art%3A10.1038%2Fs41592-025-02772-6/MediaObjects/41592_2025_2772_MOESM4_ESM.xlsx Source metadata	Source Data Fig. 2, Ahlmann-Eltze et al. 2025 Original source ↗ No field-specific location recorded Version: Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article Retrieved: 2026-10-11T00:48:13Z	catalogued No individual claim review recorded Source has a recorded evidence concern. Consult its source page before using the claim. Audit details Field: `attributes.url` Source artifact SHA-256: `2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35` Hash scope: Hash scope not separately documented; inspect source record

How to read the evidence

Sources and history

Release 2026-10-22-1b7cf027e4f6 · Record review: source checked

Authors' Source Data workbook for Fig. 2, published by the journal with the article.

Release history

No supporting source is linked yet.

Read original source

Download this release (gzip)

Technical metadata and extraction receipts

Stable ID: perturbation-ahlmann-20261011-source-data-fig2

areas: cells-tissues; single-cell
contexts: research
url: https://static-content.springer.com/esm/art%3A10.1038%2Fs41592-025-02772-6/MediaObjects/41592_2025_2772_MOESM4_ESM.xlsx
version: Source Data Fig. 2 (41592_2025_2772_MOESM4_ESM.xlsx) as published with the article
retrieved at: 2026-10-11T00:48:13Z
artifact sha256: 2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35
doi: 10.1038/s41592-025-02772-6
publication status: peer_reviewed
licence: CC-BY-4.0
media type: application/vnd.openxmlformats-officedocument.spreadsheetml.sheet
venue: Nature Methods
year: 2025
source locator: Article 'Data availability' and the 'Source Data Fig. 2' link under the figure; the article licence covers its supplementary material
scope note: 1837063 bytes. Each row of sheet "Panel A" is one held-out perturbation in one test-training split (column seed), labelled test or val in column train.
evidence concerns: source id: perturbation-ahlmann-20261011-source-data-fig2; message: The Fig. 2 legend gives "134, 210 and 24 unseen single perturbations across two test-training splits". In Panel A these are the split 1 counts only (Replogle K562 134, Replogle RPE1 210, Adamson 24 rows per method); split 2 has 128, 203 and 24, so the two-split totals are 262, 413 and 48. The Fig. 2a means recorded here use all Panel A rows; whether the plotted red lines use both splits is not stated beyond the sheet itself. Reviewer check 2026-10-11: using split 1 rows only would change the per-model means by up to 0.61 (L2 distance) and 0.06 (Pearson delta), and would change the model with the lowest mean L2 distance on Adamson and Replogle RPE1, so the choice is material.; source locator: Fig. 2 legend versus Source Data Fig. 2, sheet "Panel A"; artifact sha256: 2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35; reviewed at: 2026-10-11T00:00:00Z; review method: ai-assisted-source-review; source id: perturbation-ahlmann-20261011-source-data-fig2; message: Methods restricts the single-perturbation analysis to perturbations predicted by all models: 73 for Adamson, 398 for Replogle K562 and 629 for Replogle RPE1. No sheet of Source Data Fig. 2 has these counts (Panel A: 48, 262 and 413 per method over two splits; Panel C: 20, 214 and 224 test or val rows per method).; source locator: Methods 'Single perturbation benchmark setup' paragraph 4 versus Source Data Fig. 2, sheets "Panel A" and "Panel C"; artifact sha256: 2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35; reviewed at: 2026-10-11T00:00:00Z; review method: ai-assisted-source-review; source id: perturbation-ahlmann-20261011-source-data-fig2; message: The test and val rows in Panel C (20, 214 and 224 per method for Adamson, Replogle K562 and Replogle RPE1) match neither the Fig. 2 legend counts nor the Methods counts. Fig. 2c was not extracted.; source locator: Source Data Fig. 2, sheet "Panel C" versus the Fig. 2 legend and Methods; artifact sha256: 2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35; reviewed at: 2026-10-11T00:00:00Z; review method: ai-assisted-source-review; source id: perturbation-ahlmann-20261011-source-data-fig2; message: The main text says CPA was not included in the single-perturbation benchmark, but method "cpa" has rows in Panel C. Fig. 2c was not extracted.; source locator: Main text paragraph 16 versus Source Data Fig. 2, sheet "Panel C"; artifact sha256: 2ccfa7239c195f329e1632b51510adebfd493e0a5a7427a0f012e078e0e59c35; reviewed at: 2026-10-11T00:00:00Z; review method: ai-assisted-source-review

Related records

Suggest a correction