A Novel Multi-head Algorithm to Discover all Tandem Repeats in One Scan of DNA Sequences

Abstract

Abstract Tandem repeats (TRs) are subsequences of DNA or any genomic sequence composed of many consecutive repeats of a pattern in the same direction. TRs form about three percent of human DNA. Tandem repeats are extremely unstable and highly vulnerable to mutations. Mutated TRs can cause several diseases, such as neurodegeneration and ovarian insufficiency. The tandem repeats are divided into various classes, and consequently, computer specialists have developed many software packages, each of which is usually able to detect a class of consecutive repeats. The boundaries of these classes of tandem repeats are blurred, and it is often necessary to use more than one software tool to detect all TRs of a range of pattern lengths. Besides, if a TR with a specific pattern length is of interest, the complete software has to be run, and the output has to be manually searched. In this research, a single software package is developed to discover TRs of all repeating pattern lengths. Users are allowed to specify the range of pattern lengths of interest as input to the algorithm. The Multi-head Reader Arm (MRA) Algorithm is based on the idea of a multi-head reader arm moving on a given genomic sequence from beginning to end. Each pair of heads is composed of the base head and one of the other heads and it is responsible for finding all TRs of a specific pattern length. The innovative MRA has the ability to discover both exact and inexact TRs of all pattern lengths. Compared to the existing state-of-the-art TR detection software, we have demonstrated that MRA is superior with respect to simplicity, accuracy, computational time, and space requirement.