Figure 2

A dramatic increase of nucleotide mismatch calling from a two-round unique mapping approach. (A) Multiple mapped reads from RNA-seq data of human colon tissue were uniquely aligned to chr2: 37,328,032 -37,328,129 of the hg19 genome with the number of mismatches shown on the right. The predicted editing sites are highlighted with arrows. Reads with 4 and 5 mismatches (bold on right) were identified with the split and 2nd-round mapping approach and would have been missed with the default mapping. (B) The two-round mapping approach achieved a significant increase of mapped mismatches (bottom panel) and subtle changes of mapped reads (top panel). The 1st-round mapping, light grey bars; the split and 2nd-round mapping, dark grey bars.