Pathogenic Gene Detection For Rare Disease And Cancer Based On Alignment Strategy

Genomic variation is widespread in human genomes,which is the fundamental difference between individuals in a population.There are many types of variations in human genomes,including single nucleotide variation,small insertions and deletions,and large structural variations,such as copy number variation,inversion,and translocation.Accurate identification of genomic variations play a key role for the discovery of pathogenic genes and selection of treatments for patients with diverse diseases.This study focuses on the genomic variations,ranging from single nucleotide base to hundreds of million bases,and the interpretation of their associations with diverse diseases.Here,we propose different kinds of strategies for data analysis of whole genome sequencing data about rare disease,familial cancer,and common cancer.For rare diseases with atypical phenotypes,we perform whole genome sequencing on both affected and unaffected family members.With the assumption of inheritance pattern and pathogenic scores for the candidate variations,we identify the potential causative gene,LYST,which offers the direction for further experimental validation.Additionally,genotype-phenotype mapping also support the association between the pathogenic variations and the phenotypes.To decipher the causative genes for hepatoblastoma,we also perform whole genome sequencing on all the family members with two affected offspring.To discover the driver genes for hepato-carcinogenesis,cancer predisposition genes and genes with Mendelian inheritance are selected for pathogenicity analysis,and candidate pathogenic genes APC and WAS are successfully identified.Moreover,gene expression profiles are used to analyze the impact of variations on the expression of downstream genes or pathways.In addition,as currently popular tools for structural variation(SV)detection have their own shortcomings,we present a SV detection tool,BioSV,which can be used for detecting both germline and somatic SVs.The comparison between BioSV and currently popular tools proves BioSV to be higher sensitivity and precision.Particularly,BioSV can predict the SV breakpoints at high resolution.At the end,we identify the somatic SVs of hepatocellular carcinoma(HCC)based on the tumor and non-rumor paired whole genome sequencing data.Integrative analysis of somatic SVs and gene expression profiles further pinpoint some potential driver genes for the tumorigenesis of HCC.