Minimal gene set discovery in single-cell mRNA-seq datasets with ActiveSVM

Abstract Sequencing costs currently prohibit the application of single-cell mRNA-seq to many biological and clinical analyses. Targeted mRNA-sequencing reduces sequencing by profiling reduced gene sets that capture information with a minimal number genes. Here we introduce an active learning method identifies but highly informative enable identification cell types, physiological states genetic perturbations in data using small Our feature selection procedure generates from employing support vector machine (ActiveSVM) classifier. We demonstrate ActiveSVM ~90% cell-type classification accuracy across, for example, atlas disease-characterization datasets. The discovery should reductions measurements necessary tests, therapeutic screens.