Fig. 2

FINDER implements changepoint analysis of read coverages to modify existing gene models and/or generate new ones. Changepoint analysis is a statistical technique to assess alterations in trends over time. The same approach has been used to analyze read coverage patterns of a genome, where the data is distributed spatially. a Two Arabidopsis thaliana genes AT1G42960.1 and AT1G42970.1 are present within 50 base pairs of each other on the positive strand. Reads originating from the end exons of either genes bleed into each other resulting in PsiCLASS to merge the two gene models. Changepoint analysis recognizes the fall the read coverage and reports a position within the exon where the trough exists. This information is used to split up the gene models. b A similar issue exists with closely spaced genes residing on opposite strands. The end exons (highlighted with a red box) for a transcript extend up to the nearest intron of the adjacent transcript. Changepoint analysis is used to determine the actual end/start of transcript based on the read coverage. (Screenshot obtained from Integrative Genomics Viewer and figure generated using Microsoft PowerPoint v16.47)