A systematic comparison of single-cell perturbation response prediction models
Science Advances · 11.09.2026 · 12 (37)
DOI: 10.1126/sciadv.aed3414 ↗Kurz gefasst
Vergleichsstudie in der Zeitschrift Science Advances, erschienen am 11.09.2026. Verfasst von Lanxiang Li und 13 weiteren Autor:innen. Beteiligte Einrichtung: GMU-GIBH Joint School of Life Sciences, Guangzhou Medical University, Guangzhou, China. Hauptthemen laut MEDLINE: Single-Cell Analysis, Gene Expression Regulation, Computational Biology. Die Arbeit wurde bisher 3-mal zitiert (Crossref) und nennt 63 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: Major Project of Guangzhou National Laboratory.
Abstract (Originalsprache)
Predicting single-cell transcriptional responses to perturbations is central to dissecting gene regulation and accelerating therapeutic design, yet the field lacks a rigorous, task-spanning assessment of model behavior. We present a large-scale benchmark of 13 representative methods and baselines across 25 datasets spanning diverse perturbation modalities and species, including two primary immune-cell drug-response resources. We evaluated three core tasks—generalization to unseen single-gene perturbations, prediction of combinatorial interactions, and transfer across cell types—using 24 metrics covering expression-level accuracy, relative changes, differential expression (DE) recovery, and distributional similarity. Across tasks, performance depended strongly on perturbation effect size and evaluation perspective: Expression-level agreement was the highest for small-effect perturbations resembling controls, whereas delta- and DE-based metrics improved with larger effects, providing clearer signals. Models shared a conservative bias, with fine-tuned foundation models compressing variance and underestimating synergistic effects in combinations. PerturbNet showed superior recovery of DE signatures in Tasks 1 and 2, while no method consistently generalized across cell types in Task 3, where biological consistency dominated outcomes. This benchmark establishes current methodological limits, clarifies that different metrics probe distinct biological signals rather than redundant summaries of the same prediction problem, and provides a foundation for developing virtual-cell models that more faithfully capture heterogeneous perturbation responses.
Schlagwörter
MeSH-Schlagwörter (MEDLINE)
Autor:innen und Institutionen
- GMU-GIBH Joint School of Life Sciences, Guangzhou Medical University, Guangzhou, China.
- Guangzhou National Laboratory, Guangzhou, China.
- The University of Hong Kong, Hong Kong, China.
- School of Life Sciences, Westlake University, Hangzhou, China.
- Research Center for Industries of the Future, Westlake University, Hangzhou, China.
- Westlake Laboratory of Life Sciences and Biomedicine, Hangzhou, China.
ORCID-Profile: Lanxiang Li, Yue You, Yunlin Fu, Shihong Lu, Ye Cao, Wenle Ren, Shuangjia Zheng, Xiaodong Liu, Luyi Tian
Förderung
- Major Project of Guangzhou National Laboratory
Zitieren
APA
Li, L., You, Y., Fu, Y., Liao, W., Fan, X., Lu, S., Cao, Y., Li, B., Ren, W., Kong, J., Zheng, S., Chen, J., Liu, X., & Tian, L. (2026). A systematic comparison of single-cell perturbation response prediction models. Science Advances, 12(37). https://doi.org/10.1126/sciadv.aed3414
Vancouver
Li L, You Y, Fu Y, Liao W, Fan X, Lu S, et al. A systematic comparison of single-cell perturbation response prediction models. Science Advances. 2026;12(37). doi:10.1126/sciadv.aed3414
BibTeX
@article{Li2026,
author = {Li, Lanxiang and You, Yue and Fu, Yunlin and Liao, Wenyu and Fan, Xueying and Lu, Shihong and Cao, Ye and Li, Bo and Ren, Wenle and Kong, Jiaming and Zheng, Shuangjia and Chen, Jizheng and Liu, Xiaodong and Tian, Luyi},
title = {A systematic comparison of single-cell perturbation response prediction models},
journal = {Science Advances},
year = {2026},
volume = {12},
number = {37},
doi = {10.1126/sciadv.aed3414}
}
EUVIMED