Fig. 5

Global analysis of conserved motifs in 3′UTRs with LncLOOM. a Number of genes with various numbers of ortholog sequences that had no significant alignment to their human sequence (black) or to their mouse, dog, and chicken sequences (gray). b Distribution of combinations of unique k-mers conserved in the indicated number of sequences that did not align to the human 3′UTR sequence. c Quantification of the total number of unique k-mers (pink) and their total instances (dark red) that LncLOOM identified per species. The total number of broadly conserved miRNA binding sites is shown in green, and the number of unique k-mers that correspond to these sites in yellow. The number of genes that contained any k-mer is shown in gray, and the number of genes that contained at least one k-mer that corresponds to a miRNA site is shown in black. d Top: Distribution of unique k-mers that were identified in the first sequence non-alignable to human in multiple genes (gray). The number of k-mers detected in an invertebrate species in at least one gene is shown in black. Bottom: Unique k-mers common to at least 50 genes and detected in an invertebrate sequence. k-mers that resemble an ARE are colored red, those resembling a PAS are blue, and those resembling a PRE are green. e Comparison of genes that contained broadly conserved miRNA binding sites detected by LncLOOM and TargetScan in the human sequences of genes analyzed. f Number of broadly conserved miRNA bindings detected by LncLOOM per number of non-alignable sequences; the percentage of genes with a miRNA site detected per number of non-alignable layers (black) and the number of unique k-mers corresponding to the miRNA binding sites (yellow). g Top: Broadly conserved miRNA binding sites predicted by LncLOOM in human sequences. Sites predicted by TargetScan and recovered by LncLOOM are shown in red, and new sites (not previously predicted by TargetScan) in blue. Bottom: The conservation of these sites per number of species. h Comparison of the fractions of genes with at least one miRNA site detected in the indicated species by TargetScan and LncLOOM. Only sites found in TargetScanHuman were used. i Percentage of genes that contain a miRNA site detected by LncLOOM per number of non-alignable sequences: (red) miRNA sites that were previously predicted by TargetScan in the human sequence and recovered by LncLOOM in additional sequences, that were not part of the MSA used by TargetScan; (blue) new miRNA sites predicted by LncLOOM but not previously predicted by TargetScan in the human sequences