Error Correction and Clustering Algorithms for Next Generation Sequencing

Next generation sequencing (NGS) revolutionized genomic data generation by enabling high-throughput parallel sequencing, making large scale genomic data analysis a crucial task. To improve NGS data quality, we developed an efficient algorithm that uses a flexible read decomposition method to improve accuracy of error correction. We further proposed a statistical framework to differentiate infrequently observed sub reads from sequencing errors in the prevalence of genomic repeats. To enable the analysis of microbial organism composition in environmental samples, we developed a parallel solution for metagenomic sequence clustering integrating sketching, quasi-clique enumeration and MapReduce techniques.