Fig. 5
From: Retained introns in long RNA-seq reads are not reliably detected in sample-matched short reads
Short-read tool performance across six genes with experimentally validated RIs. Comparison of short-read tool called RIs with introns detected in long-read data are shown as a pair of matrices for each of six genes (AP1G2, LBR, SRSF7, IGSF8, FAHD2A, and FAHD2B). The rows in each matrix correspond to the results from each of eight short-read tools (from top to bottom: 1, IntEREst; 2, iREAD; 3, IRFinder-S; 4, superintronic; 5, KMA; 6, rMATS; 7, MAJIQ; 8, SUPPA2) applied to either HX1 (top) or iPSC (bottom) data; columns correspond to all introns found across all annotated transcript isoforms of the indicated gene, ordered by left and then right genomic coordinates. Each cell in the matrix depicts the presence or absence of an intron in short-read and/or long-read data as a TP (green), FN (blue), FP (pink), and TN (peach) assessment; white boxes indicate introns found only in transcripts with \(< 5\) assigned long reads. Black outlines indicate the experimentally validated RI(s) in each gene