AUTHOR=Baudry Lyam , Foutel-Rodier Théo , Thierry Agnès , Koszul Romain , Marbouty Martial 

TITLE=MetaTOR: A Computational Pipeline to Recover High-Quality Metagenomic Bins From Mammalian Gut Proximity-Ligation (meta3C) Libraries

JOURNAL=Frontiers in Genetics

VOLUME=Volume 10 - 2019

YEAR=2019

URL=https://www.frontiersin.org/journals/genetics/articles/10.3389/fgene.2019.00753

DOI=10.3389/fgene.2019.00753

ISSN=1664-8021

ABSTRACT=Characterizing the full genomic structure of complex microbial communities is a key step towards the understanding of their diversity, dynamics and evolution. These investigations are typically done through the analysis of millions of short DNA sequences directly extracted from the environment. Computational tools exploiting these metagenomics data display intrinsic limitations or constraints, such as assumptions regarding the genomic content of the genomes being investigated, and/or the need for multiple samples to accurately bin the interleaved metagenomic sequences according to their covariant characteristics. Here we present MetaTOR, an open-source and transparent computational solution that exploits meta3C, i.e. proximity ligation experiments (3C, Hi-C) performed on metagenomic samples, to bin the resulting sequencing reads into individual genomes according to their 3D contact frequencies. MetaTOR was applied on a combination of 20 newly generated meta3C libraries of mice gut microbiote sampled over time. We quantified the ability of the program to recover high-quality metagenomics-assembled metagenomes (MAGs) from metagenomics assemblies generated directly from the meta3C libraries. Whereas 16 MAGs are identified in the 148Mb assembly generated using a single meta3C library, MetaTOR identifies 122 MAGs in the 763Mb assembly generated from the merged 20 meta3C libraries, corresponding to a ~40% increase compared to MAGs recovered using current, state-of-the-art hybrid binning programs. Overall, the completion and contamination of meta3C bins were also improved. These results underline the potential of meta3C (and 3C based approaches) in metagenomics projects.