Statistical approaches to detecting and analyzing tandem repeats in genomic sequences

Abstract

Tandem repeats (TRs) are frequently observed in genomes across all domains of life. Evidence suggests that some TRs are crucial for proteins with fundamental biological functions and can be associated with virulence, resistance, and infectious/neurodegenerative diseases. Genome-scale systematic studies of TRs have the potential to unveil core mechanisms governing TR evolution and TR roles in shaping genomes. However, TR-related studies are often non-trivial due to heterogeneous and sometimes fast evolving TR regions. In this review, we discuss these intricacies and their consequences. We present our recent contributions to computational and statistical approaches for TR significance testing, sequence profile-based TR annotation, TR-aware sequence alignment, phylogenetic analyses of TR unit number and order, and TR benchmarks. Importantly, all these methods explicitly rely on the evolutionary definition of a tandem repeat as a sequence of adjacent repeat units stemming from a common ancestor. The discussed work has a focus on protein TRs, yet is generally applicable to nucleic acid TRs, sharing similar features.

Keywords: molecular evolution; protein domain; sequence profile model; tandem repeat annotation; tandem repeats.

Figure 1

Tandem repeats in genomic sequences. (A) An example TR with three units and the corresponding MSA of its units. (B) Different parts of a TR motif (R = right and L = left) have different histories after a single duplication with shifted TR units. Shown are these duplication histories as phylogenies of the right and left parts of the TR motif. (C) Five scenarios of overlapping and non-overlapping TR annotations.

Figure 2

Overview of a generic TR annotation workflow.