Skip to main content
Fig. 11 | BMC Bioinformatics

Fig. 11

From: ASGAL: aligning RNA-Seq data to a splicing graph to detect novel alternative splicing events

Fig. 11

Examples of splicing event misclassification. The figure depicts three situations in which ASGAL may detect a false positive event: an alternative donor site in (a), an alternative acceptor site in (b), and an intron retention in (c). The black arrows represent an annotated intron (in the first two cases), whereas the red dotted lines represent the novel intron supported by the input sample with respect to the annotation \(\mathcal {S}_{G}^{\star }\). In cases (a) and (b), the novel event induces an alternative 5’ and an alternative 3’ splice site (respectively) with respect to the intron in the annotation, and in case (c) the novel intron is inside an already annotated exon. On the assumption that the reads come from a hypothetical novel transcript T1, then ASGAL outputs a true positive event. Indeed, all the events refer to an annotated exon, thus one (the start of the 5’ exon in case (a) and the end of the 3’ exon in case (b)) or two splice sites (the start of the 5’ exon and the end of the 3’ exon in case (c)) involved in the predicted event are already annotated in \(\mathcal {S}_{G}^{\star }\). By contrast, if the aligned reads come from a hypothetical novel transcript T2, then ASGAL produces a false positive event: it outputs the events with respect to the annotated exon but the true events refer to a novel exon having both splice sites not annotated in \(\mathcal {S}^{\star }_{G}\)

Back to article page