Utilizing de Bruijn graph of metagenome assembly for metatranscriptome analysis

TL;DR

This paper introduces TAG, a novel method that uses de Bruijn graph-based metagenome assemblies to improve metatranscriptome assembly, enabling more comprehensive transcript reconstruction from microbial communities.

Contribution

The paper presents a new read mapping algorithm onto de Bruijn graphs and applies it to enhance metatranscriptome assembly using metagenome graphs as references.

Findings

TAG assembles more transcripts than traditional methods.

The approach captures transcripts missed by fragmented references.

The method is efficient and open source.

Abstract

Metagenomics research has accelerated the studies of microbial organisms, providing insights into the composition and potential functionality of various microbial communities. Metatranscriptomics (studies of the transcripts from a mixture of microbial species) and other meta-omics approaches hold even greater promise for providing additional insights into functional and regulatory characteristics of the microbial communities. Current metatranscriptomics projects are often carried out without matched metagenomic datasets (of the same microbial communities). For the projects that produce both metatranscriptomic and metagenomic datasets, their analyses are often not integrated. Metagenome assemblies are far from perfect, partially explaining why metagenome assemblies are not used for the analysis of metatranscriptomic datasets. Here we report a reads mapping algorithm for mapping of short reads onto a de Bruijn graph of assemblies. A hash table of junction k-mers (k-mers spanning branching structures in the de Bruijn graph) is used to facilitate fast mapping of reads to the graph. We developed an application of this mapping algorithm: a reference based approach to metatranscriptome assembly using graphs of metagenome assembly as the reference. Our results show that this new approach (called TAG) helps to assemble substantially more transcripts that otherwise would have been missed or truncated because of the fragmented nature of the reference metagenome. TAG was implemented in C++ and has been tested extensively on the linux platform. It is available for download as open source at http://omics.informatics.indiana.edu/TAG.