Bioinformatics Seminar
Time:
Venue: Na
5 July 2016
NaExactSNP : Efficient and accurate SNP calling via utilizing background count data
Yang LiaoWEHI Bioinformatics
The next-generation sequencing technologies provide an unprecedented opportunity to identify genomic mutations ; such as single nucleotide polymorphism (SNP) ; at the highest resolution. Numerous algorithms have been proposed to detect SNPs in sequence data in the past decade ; however most of the them suffer from high computational cost due to sophisticated algorithm design or computing-intensive pre-processing of the data or both.
ExactSNP is a SNP calling algorithm we have recently developed ; which uses regions flanking each candidate SNP location to detect mapping and sequencing errors and then performs Fisher's Exact test to call SNPs that have a signal stronger than the background noise. The background noise is measured by counting the number of reads mapping to the regions flanking a candidate SNP. In this talk ; I will present some latest development on ExactSNP ; in particularly on calling SNP from high sequencing-depth data. I will talk about an efficient asymptotic algorithm that we recently implemented in exactSNP for testing the significance of SNP calls in such data and I will present evaluation results showing the speed advantage of ExactSNP compared to other tools.
;;