Alternative splicing analysis

Analyzing alternative splicing (AS) in the presence of SNP within a transcript can be critical as site mutation can affect splice site recognition, leading to altered splicing pattern for intance. Hence, splicing alterations can be associated with the production of aberrant protein isoforms which potentially affect cellular function. In contrast, mutational effect of a SNP can be mitigated or overlooked by AS events, where different splice patterns can compensate for or bypass mutational impact, maintaining normal protein function or expression levels. Here, we assess AS patterns using two complementary approaches, generalized linear models implemented in the R package DEXseq [63] and Multivariate Analysis of Transcript Splicing implemented in the program rMATS [64].

For DEXseq, low expressed exons were discared, keeping only thoses where at least ten reads were aligned to a minimum of ten samples. Normalization of exon read count was performed using DESeq2 algorithm as suggested by DEXseq manual recommendation. Differential exon usage was assessed based on an FDR threshold of 0.01. While DEXseq allow only detection of differential exon usage, rMATS identifies and quantifies the major types of alternative splicing patterns, including skipped exons, alternative 5’ and 3’ splice sites, mutually exclusive exons, and retained introns. Here rMATS was performed using STAR sorted-BAM files using a pre-filtering step to retain only reads that were mapped to the scaffold43000 (113.025 kbp). False discovery rate < 0.01 and p-value < 0.05 were fixed as level of significance.