rewirebio.iobenchmarks
← Browse all sources
Source

Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

bioRxiv preprint version 1, posted 2026-07-24 · Preprint

DOI: 10.64898/2026.07.20.739620

What this source contributed

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

Results

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

Model or configurationMetricValueDatasetEvaluation
1-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q1 (0–0.017); quantile bins per cell line, averaged over four cell lines)99.07% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q2 (0.017–0.037); quantile bins per cell line, averaged over four cell lines)98.94% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q3 (0.037–0.065); quantile bins per cell line, averaged over four cell lines)98.60% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q4 (0.065–0.116); quantile bins per cell line, averaged over four cell lines)97.51% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q5 (>0.116); quantile bins per cell line, averaged over four cell lines)83.67% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))2.56% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))2.76% percentHepG2 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)1-NN on Sun et al. 2026 cross-cell-line DE classification, HepG2
1-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))3.18% percentJurkat essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)1-NN on Sun et al. 2026 cross-cell-line DE classification, Jurkat
1-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))2.13% percentK562 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)1-NN on Sun et al. 2026 cross-cell-line DE classification, K562
1-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))2.16% percentRPE1 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)1-NN on Sun et al. 2026 cross-cell-line DE classification, RPE1
1-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q1 (0–0.016); quantile bins per cell line, averaged over four cell lines)98.76% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q2 (0.016–0.034); quantile bins per cell line, averaged over four cell lines)98.62% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q3 (0.034–0.060); quantile bins per cell line, averaged over four cell lines)98.22% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q4 (0.060–0.108); quantile bins per cell line, averaged over four cell lines)97.02% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q5 (>0.108); quantile bins per cell line, averaged over four cell lines)84.65% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))3.39% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)1-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
1-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))2.98% percentHepG2 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)1-NN on Sun et al. 2026 within-cell-line DE classification, HepG2
1-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))1.69% percentJurkat essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)1-NN on Sun et al. 2026 within-cell-line DE classification, Jurkat
1-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))4.27% percentK562 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)1-NN on Sun et al. 2026 within-cell-line DE classification, K562
1-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))4.63% percentRPE1 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)1-NN on Sun et al. 2026 within-cell-line DE classification, RPE1
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q1 (0–0.017); quantile bins per cell line, averaged over four cell lines)99.86% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q2 (0.017–0.037); quantile bins per cell line, averaged over four cell lines)99.83% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q3 (0.037–0.065); quantile bins per cell line, averaged over four cell lines)99.72% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q4 (0.065–0.116); quantile bins per cell line, averaged over four cell lines)99.12% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q5 (>0.116); quantile bins per cell line, averaged over four cell lines)84.38% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))0.58% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))0.47% percentHepG2 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)10-NN on Sun et al. 2026 cross-cell-line DE classification, HepG2
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))0.73% percentJurkat essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)10-NN on Sun et al. 2026 cross-cell-line DE classification, Jurkat
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))0.63% percentK562 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)10-NN on Sun et al. 2026 cross-cell-line DE classification, K562
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))0.49% percentRPE1 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)10-NN on Sun et al. 2026 cross-cell-line DE classification, RPE1
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q1 (0–0.016); quantile bins per cell line, averaged over four cell lines)99.87% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q2 (0.016–0.034); quantile bins per cell line, averaged over four cell lines)99.84% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q3 (0.034–0.060); quantile bins per cell line, averaged over four cell lines)99.75% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q4 (0.060–0.108); quantile bins per cell line, averaged over four cell lines)99.27% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q5 (>0.108); quantile bins per cell line, averaged over four cell lines)86.90% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))0.91% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)10-NN on Sun et al. 2026 within-cell-line DE classification, four cell lines
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))0.86% percentHepG2 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)10-NN on Sun et al. 2026 within-cell-line DE classification, HepG2
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))0.28% percentJurkat essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)10-NN on Sun et al. 2026 within-cell-line DE classification, Jurkat
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))0.96% percentK562 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)10-NN on Sun et al. 2026 within-cell-line DE classification, K562
Weighted 10-NN on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))1.55% percentRPE1 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)10-NN on Sun et al. 2026 within-cell-line DE classification, RPE1
Per-gene L2 logistic regression on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q1 (0–0.017); quantile bins per cell line, averaged over four cell lines)89.03% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)Logistic regression on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Per-gene L2 logistic regression on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q2 (0.017–0.037); quantile bins per cell line, averaged over four cell lines)88.30% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)Logistic regression on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Per-gene L2 logistic regression on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q3 (0.037–0.065); quantile bins per cell line, averaged over four cell lines)86.50% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)Logistic regression on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Per-gene L2 logistic regression on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q4 (0.065–0.116); quantile bins per cell line, averaged over four cell lines)82.65% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)Logistic regression on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Per-gene L2 logistic regression on GenePT embeddings (Sun et al. 2026)Accuracy (3-class (non-DE, up, down) accuracy within true |log2 FC| quantile Q5 (>0.116); quantile bins per cell line, averaged over four cell lines)68.43% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)Logistic regression on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Per-gene L2 logistic regression on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))16.61% percentK562, RPE1, HepG2 and Jurkat test sets (Sun et al. 2026)Logistic regression on Sun et al. 2026 cross-cell-line DE classification, four cell lines
Per-gene L2 logistic regression on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))18.90% percentHepG2 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)Logistic regression on Sun et al. 2026 cross-cell-line DE classification, HepG2
Per-gene L2 logistic regression on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))18.03% percentJurkat essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)Logistic regression on Sun et al. 2026 cross-cell-line DE classification, Jurkat
Per-gene L2 logistic regression on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))12.91% percentK562 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)Logistic regression on Sun et al. 2026 cross-cell-line DE classification, K562
Per-gene L2 logistic regression on GenePT embeddings (Sun et al. 2026)Proportion (share of perturbation-gene pairs predicted as differentially expressed (compare with the printed true DE prevalence))16.62% percentRPE1 essential-gene Perturb-seq, 40% test perturbations (Sun et al. 2026)Logistic regression on Sun et al. 2026 cross-cell-line DE classification, RPE1

Claims

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

Models, benchmarks and other records

Configurations (10)

Benchmarks (1)

Protocols (2)

Dataset subsets (5)

Evaluations (85)

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.

25 evidence rows matching the loaded filters

Claims, original sources and review scope · Release 2026-10-23-5347695500ed
Property and statementOriginal source and locationReview and provenance
Artifact SHA-256
fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf
Source metadata		Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

Original source ↗

No field-specific location recorded

Version: bioRxiv preprint version 1, posted 2026-07-24
Retrieved: 2026-10-11T21:35:10Z

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: fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf

Hash scope: Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.

Inspected artifact

Artifact url
https://www.biorxiv.org/content/10.64898/2026.07.20.739620v1.full.pdf
Source metadata		Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

Original source ↗

No field-specific location recorded

Version: bioRxiv preprint version 1, posted 2026-07-24
Retrieved: 2026-10-11T21:35:10Z

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_url

Source artifact SHA-256: fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf

Hash scope: Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.

Inspected artifact

Doi
10.64898/2026.07.20.739620
Source metadata		Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

Original source ↗

No field-specific location recorded

Version: bioRxiv preprint version 1, posted 2026-07-24
Retrieved: 2026-10-11T21:35:10Z

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: fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf

Hash scope: Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.

Inspected artifact

Evidence concerns
1 values
  • source id
    lit-cells-networks-20261011-sun2026de-source-sun2026
    message
    Methods describe results on both the full data and a PerturbQA-downsampled subset ('We annotate our results to distinguish between these two settings'), but the captions of Tables 1-4 do not say which setting they report. The printed true DE prevalence (1.75-4.41%) matches the full data's 95-98% non-DE share (Table 5), not the downsampled set; this is an inference, not a printed statement.
    source locator
    Section 3.2 'Full vs. Curated Dataset'; Tables 1-5
    artifact sha256
    fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf
    review date
    2026-10-11
Source metadata		Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

Original source ↗

No field-specific location recorded

Version: bioRxiv preprint version 1, posted 2026-07-24
Retrieved: 2026-10-11T21:35:10Z

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: fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf

Hash scope: Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.

Inspected artifact

Hash scope
Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.
Source metadata		Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

Original source ↗

No field-specific location recorded

Version: bioRxiv preprint version 1, posted 2026-07-24
Retrieved: 2026-10-11T21:35:10Z

catalogued

No individual claim review recorded

Source has a recorded evidence concern. Consult its source page before using the claim.

Audit details

Field: attributes.hash_scope

Source artifact SHA-256: fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf

Hash scope: Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.

Inspected artifact

Licence
CC-BY-4.0
Source metadata		Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

Original source ↗

No field-specific location recorded

Version: bioRxiv preprint version 1, posted 2026-07-24
Retrieved: 2026-10-11T21:35:10Z

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: fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf

Hash scope: Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.

Inspected artifact

Limitations
3 values
  • Preprint, not peer reviewed.
  • Single train/test split (60/40 by perturbation); no spread printed in Tables 1-4.
  • The authors did not develop scVI, scGen or STATE. Co-author Ouyang Zhu developed Scouter and TranScouter (references 14 and 31), and co-author Yiqun T. Chen developed the GenePT embeddings used by the KNN, logistic-regression, VAE and scVI configurations (reference 30). The majority vote, KNN and logistic regression baselines and the scGen-based VAE adaptation are the authors' own.
Source metadata		Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

Original source ↗

No field-specific location recorded

Version: bioRxiv preprint version 1, posted 2026-07-24
Retrieved: 2026-10-11T21:35:10Z

catalogued

No individual claim review recorded

Source has a recorded evidence concern. Consult its source page before using the claim.

Audit details

Field: attributes.limitations

Source artifact SHA-256: fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf

Hash scope: Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.

Inspected artifact

Media type
application/pdf
Source metadata		Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

Original source ↗

No field-specific location recorded

Version: bioRxiv preprint version 1, posted 2026-07-24
Retrieved: 2026-10-11T21:35:10Z

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: fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf

Hash scope: Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.

Inspected artifact

Publication status
preprint
Source metadata		Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

Original source ↗

No field-specific location recorded

Version: bioRxiv preprint version 1, posted 2026-07-24
Retrieved: 2026-10-11T21:35:10Z

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: fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf

Hash scope: Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.

Inspected artifact

Retrieved at
2026-10-11T21:35:10Z
Source metadata		Beyond Expression Prediction: Benchmarking Differential Expression Classification in Single-Cell Perturbation Models

Original source ↗

No field-specific location recorded

Version: bioRxiv preprint version 1, posted 2026-07-24
Retrieved: 2026-10-11T21:35:10Z

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: fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf

Hash scope: Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.

Inspected artifact

How to read the evidence

Sources and history

Release 2026-10-23-5347695500ed · Record review: source checked

bioRxiv PDF, retrieved and hashed for the 2026-10-11 cells-spatial literature pass.

Release history

No supporting source is linked yet.

Read original source

Download this release (gzip)
Technical metadata and extraction receipts

Stable ID: lit-cells-networks-20261011-sun2026de-source-sun2026

areas
single-cell
contexts
research
url
https://doi.org/10.64898/2026.07.20.739620
artifact url
https://www.biorxiv.org/content/10.64898/2026.07.20.739620v1.full.pdf
version
bioRxiv preprint version 1, posted 2026-07-24
retrieved at
2026-10-11T21:35:10Z
artifact sha256
fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf
doi
10.64898/2026.07.20.739620
publication status
preprint
licence
CC-BY-4.0
media type
application/pdf
hash scope
Exact bytes of one download; bioRxiv serves slightly different PDF bytes per download, so the gzip copy in artifacts/ is the reference.
limitations
Preprint, not peer reviewed.; Single train/test split (60/40 by perturbation); no spread printed in Tables 1-4.; The authors did not develop scVI, scGen or STATE. Co-author Ouyang Zhu developed Scouter and TranScouter (references 14 and 31), and co-author Yiqun T. Chen developed the GenePT embeddings used by the KNN, logistic-regression, VAE and scVI configurations (reference 30). The majority vote, KNN and logistic regression baselines and the scGen-based VAE adaptation are the authors' own.
evidence concerns
source id: lit-cells-networks-20261011-sun2026de-source-sun2026; message: Methods describe results on both the full data and a PerturbQA-downsampled subset ('We annotate our results to distinguish between these two settings'), but the captions of Tables 1-4 do not say which setting they report. The printed true DE prevalence (1.75-4.41%) matches the full data's 95-98% non-DE share (Table 5), not the downsampled set; this is an inference, not a printed statement.; source locator: Section 3.2 'Full vs. Curated Dataset'; Tables 1-5; artifact sha256: fa1c6a6345c45e14b9c373dfa8aff12633ae646feece1cfa3899c3ed61f6cfdf; review date: 2026-10-11
review
method: source-hash-verification; method note: Archived gzip copy unpacked and its SHA-256 matched; a fresh bioRxiv download was rate-limited (HTTP 429).; reviewer: claude; reviewer note: Claude review agent reviewer-cn-1, independent of the extractor lit-cells-networks; no human review claimed; date: 2026-10-12
Related records

    Suggest a correction