Calling Genotypes from Bisulfite Sequencing Data

Abstract

Methylation is studied to be a most importance mechanism of epigenetics, which is significant in gene expressing, cell differentiations and human diseases. In bioinformatics, methylation analysis is quite time consuming faced with gigabytes of data. It is urgent and meaningful to enhance the efficiency of the methylation analysis software and tools. This project improved the efficiency of general methylation analysis by optimizing the basic procedures of methylation analysis, bisulfite sequencing alignment and the following genotypes calling. A new algorithm was implemented in alignment, interpreting the sequence into two patterns for mapping and lengthening the seed size to add the error tolerance of the seeds. Calling genotypes was optimized in calculating algorithms and implemented in C++. After testing, the alignment efficiency is enhanced strikingly and the aligned ratio is slightly increased. Calling genotypes is more than ten times faster than before. The two steps are run by separate software and are streamlined in a popular methylation analysis pipeline. In addition, a data visualization module is conducted in the end the pipeline to generally demonstrate the methylation analysis results.