Hi-C deconvolution of a human gut microbiome yields high-quality draft genomes and reveals plasmid-genome interactions

The assembly of high-quality genomes from mixed microbial samples is a long-standing challenge in genomics and metagenomics. Here, we describe the application of ProxiMeta, a Hi-C-based metagenomic deconvolution method, to deconvolve a human fecal metagenome. This method uses the intra-cellular proximity signal captured by Hi-C as a direct indicator of which sequences originated in the same cell, enabling culture-free de novo deconvolution of mixed genomes without any reliance on a priori information. We show that ProxiMeta deconvolution provides results of markedly high accuracy and sensitivity, yielding 50 near-complete microbial genomes (many of which are novel) from a single fecal sample, out of 252 total genome clusters. ProxiMeta outperforms traditional contig binning at high-quality genome reconstruction. ProxiMeta shows particularly good performance in constructing high-quality genomes for diverse but poorly-characterized members of the human gut. We further use ProxiMeta to reconstruct genome plasmid content and sharing of plasmids among genomes—tasks that traditional binning methods usually fail to accomplish. Our findings suggest that Hi-C-based deconvolution can be useful to a variety of applications in genomics and metagenomics.