matched for GC content and average accessibility (Supplementary Figs. 2–5; Supplementary Note 2). The raw accessibility deviations for these background peak sets are used to compute a bias corrected deviation and z-score for each annotation and cell, providing a bias-corrected differential measure describing the gain or loss of accessibility of a given genomic annotation relative to the average cell profile (see methods). These bias corrected deviations and z-scores can be used for a number of downstream analyses, including de novo clustering of cells and identification of key regulators that vary within and between different cell types. The chromVAR package includes a collection of tools for such downstream analysis, including an interactive web application for exploring the relationship between key TF motifs and clustering of cells (Supplementary Figure 6). We have also incorporated tools for generating previously-described analyses characterizing the correlation and potential cooperativity between two TF binding sites within the same regulatory element, and computing chromatin variability across regions in cis4.
