Perturbscape framework uses CRISPR Perturb-seq to map hidden regulatory architecture of complex diseases
A preprint introduces Perturbscape, a computational framework that learns gene regulatory programmes from CRISPR perturbation screens and integrates them with human genetics data to identify disease-relevant genes and pathways.
A preprint posted to bioRxiv describes Perturbscape, a computational framework designed to extract regulatory gene programmes from CRISPR Perturb-seq screens — large-scale experiments in which individual genes are disrupted by CRISPR and the resulting changes in gene expression are measured at single-cell resolution.
The authors argue that steady-state molecular profiles, such as those obtained from standard transcriptomic studies of patient tissue, incompletely capture the regulatory logic through which genetic variants influence complex disease. Perturbational data, by contrast, can reveal programmes that only become apparent when cells are subjected to defined genetic or environmental stresses.
Perturbscape combines linear and non-linear contrastive learning methods with complementary programme-learning approaches to characterise multiscale gene programmes across a screen of 15,859 perturbations — a substantial scale for this class of experiment. The framework then integrates the resulting programme landscape with human genetics data, including GWAS summary statistics and fine-mapping results, to prioritise disease-relevant genes and their downstream regulatory targets.
The preprint does not name a specific journal for peer-reviewed publication. As a preprint, the results have not yet undergone formal peer review and should be interpreted accordingly. The methodological approach is relevant to researchers working on functional genomics, post-GWAS interpretation, and the development of computational tools for gene regulatory analysis. The scale of the perturbation screen and the integration with human genetics data are notable features of the work.
Sources
Read the original reporting — these are the public sources this summary draws from.
-
Primary sourcePreprint bioRxiv (Cold Spring Harbor Laboratory) · 2026-11-04Perturbation-induced gene programs map the hidden regulatory architecture of complex diseases