Analysis Of Allele Specific Alternative Splicing Based On Single Nucleotide Variants

As the development of sequencing technology, human are slowly unveilingthe secret of human genome. The focus of human genome analysis is starting totransform from deciphering gene codes to knowing more about the function ofgenes and unearth the association between gene and disease. In a risingresearching field like bioinformatics, the study of gene expression and transcriptregulation is becoming more and more popular. In the mechanism of geneexpression and transcript regulation, alternative splicing has been playing a keyrole. Alternative splicing not only enriches the information th at genome cancontain, but also is highly related with disease. This dissertation proposed asystem that can process gene data, and estimate the expression level of differentalleles and different isoforms of genes through the analysis of single nucleotidevariants based allele specific alternative splicing.Since the alternative events are rather complicated, this dissertationproposed a simpler model to represent alternative events in a mathematics way.This new model takes advantages of the potential information in RNA-seq readsthat other models did not use to make a more accurate and stable solution. Themodel proposed in this dissertation is optimized by particle swarm optimizationalgorithm. Moreover, this dissertation proposed a data process pipe tha t cansimplify the gene data for easier use in this and other studies. In the testing phase,the data process pipe and modeling system both perform well in real data ofU87MG cells and simulation data.