“…Gene ontology enrichment analysis was performed on the top 250 genes from each component (from each side) using the enrichGO function from the clusterProfiler R package in which p-values are calculated based on the hypergeometric distribution and corrected for testing of multiple biological process GO terms using the Benjamini-Hochberg procedure (Yu et al, 2012). Plots were created using the ggplot2 package and extensions ggrepel, ggforce, ggseqlogo, ggnewscale, ggrastr, RColorBrewer, viridis, and cowplot (Campitelli, 2019; Garnier, 2018; Neuwirth, 2014; Pedersen, 2016; Petukhov, 2018; Wagih, 2017; Wickham, 2016; Wilke, 2018). In addition the following R packages were used: data.table, Matrix (for sparse large matrix computations), biomaRt (for gene identifiers), shiny and shinycssloaders (for creating the interactive web application), ComplexHeatmap, bigmemory (for creating a low-memory shiny app), and MASS (for kernel density estimation) (Bates and Maechler, 2018; Chang et al, 2018; Dowle and Srinivasan, 2019; Durinck et al, 2005; Gu et al, 2016; Kane et al, 2013; Sali, 2017; Venables and Ripley, 2002).…”