A split-control criterion for few-shot adaptation in single-cell perturbation prediction

Preprint (not peer-reviewed): Liang et al. (Macau University of Science and Technology; Jinan University) show that conventional evaluation of few-shot adaptation in single-cell perturbation prediction shares one control estimate between calibration responses and held-out targets, which carries control error into predictions and inflates scores. Scoring against an independent control changed the rank of 40 of 52 configuration–dataset entries. They introduce SafeDelta, a split-control criterion that uses calibration data to decide whether to adapt; across six datasets, mean gated gain was non-negative under the original two-control evaluation.

Authors / organisation

Liang W-L, Yang Y-B, Zhang Q-Y, Feng J-C, Wang Q-F, Zou D, Tang J-L, Liu Y, Wang S-Y, Duan J, Bai Y, Li Y-F, Sun W-Y, Cao Y-F, He R-R.

Venue

bioRxiv (preprint, not peer-reviewed)

Date

2026-10-13

Type

preprint

Year

2026

DOI

10.64898/2026.10.05.756906

Verification

bioRxiv API details (title, 15 authors, posted 2026-10-13, v1, new results, bioinformatics, CC BY-NC-ND, Macau University of Science and Technology / Jinan University, abstract naming SafeDelta) fetched 2026-10-20; full text not read

PerturbationBenchmark & validationCritique

Open in the searchable library → · JSON