geneCover identifies a minimally redundant set of marker genes characterizing diverse cell types in the CBMC data set. To obtain the geneCover marker panel, we apply Iterative-GeneCover in Algorithm 1 with parameters: {k_t}_1:T = {50, 50, 50, 50, 50, 50, 50, 50, 50, 50}, m = 6. (A) Spearman's correlation heat map of the 50 geneCover marker genes identified by geneCover, with gene reordered by hierarchical clustering. (B) Expression matrix of geneCover markers, with same ordering as in A, in cell types. Gene expression is standardized to [0, 1] range. The color intensity represents the level of the normalized expression. (C) Same as A but for geneBasis markers. (D) Same as B but for geneBasis markers. (E) NMI versus marker panel size for different label-free methods. (F) Specificity score versus marker panel size for different label-free methods. Lower score indicates better marker specificity.