Abstract
Motivation: The low cost of DNA sequencing has accelerated research in natural product biosynthesis allowing us to rapidly link small molecules to the clusters that produce them. However, the large amount of data means that the number of putative biosynthetic gene clusters (BGCs) far exceeds our ability to experimentally characterize them. This necessitates the need for development of further tools to analyze putative BGCs to flag those of interest for further characterization.
Results: Clustertools implements a framework to aid in the characterization of putative BGCs. It does this by or-ganizing genomic information on coding sequences in a way that enables directed, hypothesis-driven queries for functional elements in close physical proximity of each other. Genomic sequence databases can be constructed in clusterTools with an interface to the NCBI Genbank and Genomes databases, or from private sequence databases. clusterTools can be used either to identify interesting BGCs from a database of putative BGCs, or on databases of genomic sequences to identify and download regions of interest in the DNA for further processing and annotation in programs such as antiSMASH. We have used clusterTools to identify putative and known biosynthetic gene clus-ters involved in bacterial polyketide alkaoloid and tetronate biosynthesis.
Availability and Implementation: Clustertools is implemented in Python and is available via the AGPL. Stand-alone versions of clusterTools are available for Macintosh, Windows, and Linux upon registration (https://goo.gl/forms/QRKTkpqiA0g31IWp1). The source-code is available at https://www.github.com/emzodls/clusterArch.
Supplementary information: A manual describing the Python toolkit that powers clusterTools, as well as the HMMs constructed for the tetronate search is available online.