Skip to main content
Fig. 3 | BMC Biology

Fig. 3

From: Evidence for widespread existence of functional novel and non-canonical human transcripts

Fig. 3

Genome-wide distribution of IACFs and CACFs. a Schematic diagram of the analytical pipeline for identification and hierarchical mapping of IACFs and CACFs to different genomic elements and regions shown in the pie charts in the panels e and f. Number of insertions in each site was compared among the 3 biological replicates of the samples subjected to a phenotypic challenge (time of growth or drug treatment) and the corresponding control to identify IACFs using FDR threshold of 25%. In the independent approach, clusters of phenotypic insertions were first identified using SICER with a FDR threshold of 1%, and those shared by at least 2 biological replicates were defined as CACFs. For more details, see Additional file 1: Fig. S1. b Venn diagram of overlap between IACFs (black) and CACFs (white). The numbers represent non-unique events where the count for the same insertion or cluster would be added for each system where it was found. c, d Ratio of real vs simulated SICER clusters (c) and CACFs (d) based on 100 simulations for one representative example of the simulation analysis for insertions depleted in one biological replicate of the K562 growth survival experiment (see Additional file 1: Fig. S2 for the complete set of simulation analyses). Error bars indicate SD. e, f Genome-wide distributions of all IACFs and CACFs across the 4 phenotypic assay systems. Each unique IACF and CACF position was assigned to only one genomic element or region as shown in a. g, h Odds ratios of enrichment of the phenotypic (depleted or enriched) and no phenotype insertions in different genomic elements. i Distribution of p values of IACFs and CACFs mapping inside or outside of the annotations defined as all elements from exons to insulators in panel a. j Conservation scores of the phenotypic vs non-phenotypic insertions based on IACFs (left) or insertions located inside or outside of the CACFs (right). Error bars indicate SD based on the 4 phenotypic assay systems. The p values were calculated with one-sided paired Student’s t test. Source data are provided in Additional file 2: Tables S4-S10

Back to article page