BWT for string collections

TL;DR

This paper surveys methods for extending the Burrows-Wheeler Transform (BWT) to collections of strings, analyzing their properties, categorizing tools, and exploring their impact on bioinformatics applications.

Contribution

It provides a comprehensive categorization of BWT variants for string collections and discusses the impact of different methods on transform properties and biological data compression.

Findings

Optimal BWT minimizes the number of runs.

Method choice affects dynamicity and compression.

Heuristics influence biological sequence compression.

Abstract

We survey the different methods used for extending the BWT to collections of strings, following largely [Cenzato and Lipt\'ak, CPM 2022, Bioinformatics 2024]. We analyze the specific aspects and combinatorial properties of the resulting BWT variants and give a categorization of publicly available tools for computing the BWT of string collections. We show how the specific method used impacts on the resulting transform, including the number of runs, and on the dynamicity of the transform with respect to adding or removing strings from the collection. We then focus on the number of runs of these BWT variants and present the optimal BWT introduced in [Cenzato et al., DCC 2023], which implements an algorithm originally proposed by [Bentley et al., ESA 2020] to minimize the number of BWT-runs. We also discuss several recent heuristics and study their impact on the compression of biological sequences. We conclude with an overview of the applications and the impact of the BWT of string collections in bioinformatics.