Calculating the Most Likely Intron Splicing Orders in S. pombe, Fruit Fly, Zebrafish, and Humans

Background Introns have been shown to be spliced in a defined order, and this order influences both alternative splicing regulation and splicing fidelity, but previous studies have only considered neighbouring introns. The detailed intron splicing order remains unknown. Results In this work, I developed a method that can calculate the intron splicing orders of all introns in each transcript. A simulation study showed that this method can accurately calculate intron splicing orders. I further applied this method to real pombe, fruit fly, zebrafish and human sequencing datasets and found that intron splicing orders change from gene to gene and that humans contain more not in-order spliced transcripts than S. pombe, fruit fly and zebrafish. In addition, I reconfirmed that the first introns in humans are spliced slower than those in S. pombe, fruit fly and zebrafish genome-widely. Both the calculated most likely orders and the method developed here are available on the web. Conclusions I developed a novel computational method to calculate the intron splicing orders and applied the method to real sequencing datasets. I obtained intron splicing orders for hundreds or thousands of genes in four organisms. I found humans contain more number of not in-order spliced transcripts. Keywords : Splicing; Intron splicing order; Most likely order; Bayesian network