Figure 1

Reconstruction of effective transcript copies by Astroid. (a) The two isoforms from which transcript fragments are randomly sampled. (b) The alignments of the sampled fragments, plotted with IGV [36]. A splice graph can be built based upon the exons and splice junctions identified from the fragment alignments. (c) Effective transcript copies assembled by Astroid. Astroid successfully reconstructs the two expressed isoforms with no false positive. (d) The distribution of fragments in the effective copies. The likelihood of each copy is assessed according to the sizes of the fragments in the copy together with the between-fragment distances. Effective transcript copies will be identified and used to measure the abundance of each isoform. Note that this example shows only transcript fragments rather than the RNA-seq reads for simplified illustration. However, our method does take paired-end reads as input.