Title
Base calling and SNP calling on next generation sequencing data
Speaker
Xinping Cui, U.C. Riverside
Abstract
Recent advances in next-generation sequencing (NGS) technology now provide the potential to detect all single nucleotide polymorphisms (SNPs) including rare ones in a genomic region. The power of NGS based SNP detection is critically dependent upon the accuracy of base calling. In this talk, I will discuss our recently developed base calling algorithm that was built on random-effect-multivariate-mixture-model and implemented by rejection controlled EM algorithm. We will demonstrate that the proposed base caller improves the precision and speed of base calling compared with currently leading methods and as a result significantly improves the power of SNP detection. A review of the available SNP calling procedures for NGS data reveals that most rely mainly on base-calling and mapping qualities as sources of error when calling SNPs. Thus errors involved in genomic sample preparation are not accounted for. In this talk, I will also discuss our new SNP caller, Genotype Model Selection (GeMS), which accounts for genomic sample preparation errors. Simulations and real data analyses indicate that GeMS has the best performance balance of sensitivity and positive predictive value among the tested SNP callers.
