A method for homopolymer length discrimination in ion torrent sequencing

Ion torrent is a novel semiconductor long sequencing platform and advantages in faster and cheaper sequencing calls, but its raw data analysis is far from perfect, especially about discrimination of homopolymer length. With Bayesian inference and reference genome information, a comprehensive method was designed to resolve such problem. Bayesian posterior probabilities of homopolymer length were firstly calculated according to distributions of raw detected voltage signals. Integrated with reference genome sequences information, the homopolymer length was eventually deduced in high precision. In the experiment, excellent performance of the proposed method was verified.