Article

Signal, Bounds, and Baselines: Principles for Evaluating Virtual Cell Perturbation Models

2026-04-22

Abstract excerpt

Foundation models and deep learning systems are increasingly proposed as core components of “virtual cells” capable of forecasting transcriptomic responses to unseen perturbations. Yet rigorous evaluation of perturbation prediction in high-dimensional gene expression space remains challenging, raising concerns of reported performances. Here, we introduce the SBB principles (Signal, Bounds, and Baselines) for evalu...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: 90a4dce4-732a-5805-97cc-4f75838d1946
DOI: 10.64898/2026.04.20.719650

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Signal, Bounds, and Baselines: Principles for Evaluating Virtual Cell Perturbation ModelsDOI 10.64898/2026.04.20.719650Read

Benchmarking a foundational cell model for post-perturbation RNAseq predictionSemantic proximity Read

Benchmarking AI Models for <i>In Silico</i> Gene Perturbation of CellsSemantic proximity Read

Deep learning-based predictions of gene perturbation effects do not yet outperform simple linear baselinesSemantic proximity Read

PertEval-scFM: Benchmarking Single-Cell Foundation Models for Perturbation Effect PredictionSemantic proximity Read

Tabular Foundation Models Are Competitive Cellular Perturbation Predictors Across Biological ScalesSemantic proximity Read

Multimodal foundation model predicts zero-shot functional perturbations and cell fate dynamicsSemantic proximity Read

geneRNIB: a living benchmark for gene regulatory network inferenceSemantic proximity Read

Learning Perturbation-specific Cell Representations for Prediction of Transcriptional Response across Cellular ContextsSemantic proximity Read

Back to basics: Observed statistics are sufficient to predict drug responsesSemantic proximity Read

Signature Recontextualization: Mapping perturbational signatures across biological contextsSemantic proximity Read

Cell-Level Virtual ScreeningSemantic proximity Read

Method Choice, Not Biology, Determines In Silico Perturbation Results: A Systematic Evaluation of Eight Methods Across Four DatasetsSemantic proximity Read