Fig. 1

Flowchart showing the various steps employed to study the difference in the extent of conservation of binding sites of RBPs across species. BED files containing binding site coordinates of human RBPs (60 files; one for each RBP) were downloaded from CLIPdb [28]. Multiple Alignment Format (MAF) files (22 files; one for each human chromosome) were downloaded from UCSC genome browser [29], which contain multiple alignments of the whole genomes of 46 vertebrate species arranged in a series of blocks. If the start and end coordinates of a binding site of an RBP from the BED file occurred within the human genome coordinates of a block in a MAF file, the block was extracted. Otherwise, the binding site was ignored. The percentage of species each binding site of an RBP was conserved in was computed from its corresponding MAF block. Repeating this procedure for all RBPs revealed the extent of conservation of binding sites for each RBP. To study the factors contributing to the differences in the extent of conservation of binding sites of RBPs, various RBP-centric and RBP-target level features were examined: A) Phylogenetic relationship of RBPs belonging to the same family, and their extent of conservation of binding sites. B) A multivariate analysis to uncover the RBP-centric features that could influence the extent of conservation of binding sites. C) The extent of conservation of binding sites depending on their location of occurrence along the length of a gene. D) Gene set enrichment to identify the phenotypes associated with an RBP’s post-transcriptional network, ranked by the percentage of species binding sites of an RBP’s target gene were conserved in