| Comparison: adaptation None: predicts the control distribution Context-only references | PerturbPFN: Probing the Limits of Synthetic Priors in Drug Perturbation Modelling (arXiv:2607.23447v1) Original source ↗ Table 2, row 'Observational' Version: arXiv v1, submitted 2026-07-26 Retrieved: 2026-11-15T20:27:40Z | not individually reviewed No individual claim review recorded author reported Audit detailsField: attributes.comparison.adaptation Source artifact SHA-256: 42191efa04b45ca04b2be38b51ea374cd9bf891efe2399d331fec9eaeab200b1 Hash scope: Exact bytes of the PDF. Three downloads on 2026-11-15 (20:22Z to 20:27Z) gave identical bytes. Inspected artifact |
|---|
| Comparison: aggregation Mean over 12 protocols per metric; overall score is the mean of the three metric means Context-only references | PerturbPFN: Probing the Limits of Synthetic Priors in Drug Perturbation Modelling (arXiv:2607.23447v1) Original source ↗ Table 2, row 'Observational' Version: arXiv v1, submitted 2026-07-26 Retrieved: 2026-11-15T20:27:40Z | not individually reviewed No individual claim review recorded author reported Audit detailsField: attributes.comparison.aggregation Source artifact SHA-256: 42191efa04b45ca04b2be38b51ea374cd9bf891efe2399d331fec9eaeab200b1 Hash scope: Exact bytes of the PDF. Three downloads on 2026-11-15 (20:22Z to 20:27Z) gave identical bytes. Inspected artifact |
|---|
| Comparison: budget No fitting Context-only references | PerturbPFN: Probing the Limits of Synthetic Priors in Drug Perturbation Modelling (arXiv:2607.23447v1) Original source ↗ Table 2, row 'Observational' Version: arXiv v1, submitted 2026-07-26 Retrieved: 2026-11-15T20:27:40Z | not individually reviewed No individual claim review recorded author reported Audit detailsField: attributes.comparison.budget Source artifact SHA-256: 42191efa04b45ca04b2be38b51ea374cd9bf891efe2399d331fec9eaeab200b1 Hash scope: Exact bytes of the PDF. Three downloads on 2026-11-15 (20:22Z to 20:27Z) gave identical bytes. Inspected artifact |
|---|
| Comparison: dataset version Sci-Plex subset preprocessed as in the GIM benchmark (Schneider et al. 2025): median-control-count normalisation, filtering, log transform, 50 marker genes per protocol chosen from training conditions Context-only references | PerturbPFN: Probing the Limits of Synthetic Priors in Drug Perturbation Modelling (arXiv:2607.23447v1) Original source ↗ Table 2, row 'Observational' Version: arXiv v1, submitted 2026-07-26 Retrieved: 2026-11-15T20:27:40Z | not individually reviewed No individual claim review recorded author reported Audit detailsField: attributes.comparison.dataset_version Source artifact SHA-256: 42191efa04b45ca04b2be38b51ea374cd9bf891efe2399d331fec9eaeab200b1 Hash scope: Exact bytes of the PDF. Three downloads on 2026-11-15 (20:22Z to 20:27Z) gave identical bytes. Inspected artifact |
|---|
| Comparison: inputs Control cells, drug identity and dose descriptors and the expression of every context condition; no drug targets Context-only references | PerturbPFN: Probing the Limits of Synthetic Priors in Drug Perturbation Modelling (arXiv:2607.23447v1) Original source ↗ Table 2, row 'Observational' Version: arXiv v1, submitted 2026-07-26 Retrieved: 2026-11-15T20:27:40Z | not individually reviewed No individual claim review recorded author reported Audit detailsField: attributes.comparison.inputs Source artifact SHA-256: 42191efa04b45ca04b2be38b51ea374cd9bf891efe2399d331fec9eaeab200b1 Hash scope: Exact bytes of the PDF. Three downloads on 2026-11-15 (20:22Z to 20:27Z) gave identical bytes. Inspected artifact |
|---|
| Comparison: metric implementation MD, Pearson r of mean vectors and debiased Sinkhorn W2 (epsilon 0.1, 200 iterations, tolerance 1e-6), each oriented higher-is-better and min-max normalised across the seven Table 2 rows within each protocol Context-only references | PerturbPFN: Probing the Limits of Synthetic Priors in Drug Perturbation Modelling (arXiv:2607.23447v1) Original source ↗ Table 2, row 'Observational' Version: arXiv v1, submitted 2026-07-26 Retrieved: 2026-11-15T20:27:40Z | not individually reviewed No individual claim review recorded author reported Audit detailsField: attributes.comparison.metric_implementation Source artifact SHA-256: 42191efa04b45ca04b2be38b51ea374cd9bf891efe2399d331fec9eaeab200b1 Hash scope: Exact bytes of the PDF. Three downloads on 2026-11-15 (20:22Z to 20:27Z) gave identical bytes. Inspected artifact |
|---|
| Comparison: population 12 protocols: A549, K562 and MCF7 by belinostat, dacinostat, givinostat and quisinostat 2HCl Context-only references | PerturbPFN: Probing the Limits of Synthetic Priors in Drug Perturbation Modelling (arXiv:2607.23447v1) Original source ↗ Table 2, row 'Observational' Version: arXiv v1, submitted 2026-07-26 Retrieved: 2026-11-15T20:27:40Z | not individually reviewed No individual claim review recorded author reported Audit detailsField: attributes.comparison.population Source artifact SHA-256: 42191efa04b45ca04b2be38b51ea374cd9bf891efe2399d331fec9eaeab200b1 Hash scope: Exact bytes of the PDF. Three downloads on 2026-11-15 (20:22Z to 20:27Z) gave identical bytes. Inspected artifact |
|---|
| Comparison: protocol id perturbpfn-20261115-protocol-sciplex-highest-dose-normalised Context-only references | PerturbPFN: Probing the Limits of Synthetic Priors in Drug Perturbation Modelling (arXiv:2607.23447v1) Original source ↗ Table 2, row 'Observational' Version: arXiv v1, submitted 2026-07-26 Retrieved: 2026-11-15T20:27:40Z | not individually reviewed No individual claim review recorded author reported Audit detailsField: attributes.comparison.protocol_id Source artifact SHA-256: 42191efa04b45ca04b2be38b51ea374cd9bf891efe2399d331fec9eaeab200b1 Hash scope: Exact bytes of the PDF. Three downloads on 2026-11-15 (20:22Z to 20:27Z) gave identical bytes. Inspected artifact |
|---|
| Comparison: split Per cell line and drug pair, the highest dose is the query; context is controls, lower doses of that drug and all doses of the other three drugs in the same cell line Context-only references | PerturbPFN: Probing the Limits of Synthetic Priors in Drug Perturbation Modelling (arXiv:2607.23447v1) Original source ↗ Table 2, row 'Observational' Version: arXiv v1, submitted 2026-07-26 Retrieved: 2026-11-15T20:27:40Z | not individually reviewed No individual claim review recorded author reported Audit detailsField: attributes.comparison.split Source artifact SHA-256: 42191efa04b45ca04b2be38b51ea374cd9bf891efe2399d331fec9eaeab200b1 Hash scope: Exact bytes of the PDF. Three downloads on 2026-11-15 (20:22Z to 20:27Z) gave identical bytes. Inspected artifact |
|---|
Limitations2 values- Run by the PerturbPFN authors, who developed the proposed method and ran every comparator.
- Min-max normalised within the seven compared rows: the value changes if a method is added or removed, so it is comparable only within Table 2.
Context-only references | PerturbPFN: Probing the Limits of Synthetic Priors in Drug Perturbation Modelling (arXiv:2607.23447v1) Original source ↗ Table 2, row 'Observational' Version: arXiv v1, submitted 2026-07-26 Retrieved: 2026-11-15T20:27:40Z | not individually reviewed No individual claim review recorded author reported Audit detailsField: attributes.limitations Source artifact SHA-256: 42191efa04b45ca04b2be38b51ea374cd9bf891efe2399d331fec9eaeab200b1 Hash scope: Exact bytes of the PDF. Three downloads on 2026-11-15 (20:22Z to 20:27Z) gave identical bytes. Inspected artifact |
|---|