Understanding the diverse responses of individual cells to the same perturbation is central to many biological and biomedical problems. Current methods, however, fall short in precisely quantifying the heterogenous perturbation responses and, more importantly, unveiling new biological insights from such heterogeneity. Here we introduce the "Perturbation Score" (PS) method, based on constrained quadratic optimization, to quantify diverse perturbation responses at a single-cell level. Applied to single-cell transcriptomes of large-scale genetic perturbation datasets (e.g., Perturb-seq), PS outperforms existing methods in quantifying partial gene perturbation responses. In addition, PS presents two major advances over existing methods. First, PS enables large-scale, single-cell resolution dosage analysis of perturbation, without the need to titrate perturbation strength. By analyzing the dosage-response patterns of over 2,000 essential genes in Perturb-seq, we identify two distinct patterns, depending on whether a moderate reduction in their expression induces strong downstream expression alterations. Second, PS identifies intrinsic and extrinsic biological determinants for perturbation responses. We demonstrate the application of PS in various contexts like T cell stimulation, latent HIV-1 expression, and pancreatic cell differentiation. Notably, PS unveiled a previously unrecognized, cell-type specific role of CCDC6 (coiled-coil domain containing 6) in guiding liver and pancreatic lineage decisions, where CCDC6 knockouts drives the endoderm cell differentiation towards liver lineage, rather than pancreatic lineage. Collectively, our PS approach provides an innovative methodology to perform dosage-to-function analysis and foster new biological discoveries from single-cell perturbation datasets.

Decoding Heterogenous Single-cell Perturbation Responses