Assembling Whole Genomes from Mixed Microbial Communities Using Hi-C

Assembly of whole genomes from next-generation sequencing is inhibited by the lack of contiguity information in short-read sequencing.  This limitation also impedes metagenome assembly, since one cannot tell which sequences originate from the same species within a population.  We have overcome these bottlenecks by adapting a chromosome conformation capture technique (Hi-C) for the deconvolution of metagenomes and the scaffolding of de novo assemblies of individual genomes.

     In modeling the 3D structure of a genome, chromosome conformation capture techniques such as Hi-C are used to measure long-range interactions of DNA molecules in physical space.  These tools employ crosslinking of chromatin in intact cells followed by intra-molecular ligation, joining DNA fragments that were physically nearby at the time of crosslink.  Subsequent deep sequencing of these DNA junctions generates a genome-wide contact probability map that allows the 3D modeling of genomic conformation within a cell.  The strong enrichment in Hi-C signal between genetically neighboring loci allows the scaffolding of entire chromosomes from fragmented draft assemblies.  Hi-C signal also preserves the cellular origin of each DNA fragment and its interacting partner, allowing for deconvolution and assembly of multi-chromosome genomes from a mixed population of organisms. 

     We have used Hi-C to scaffold whole genomes of animals, plants, fungi, as well as prokaryotes and archaea.  We have also been able to use this data to annotate functional features of microbial genomes, such as centromeres in many fungal species.  Additionally, we have applied our technology to diverse metagenomic populations such as craft beer, bacterial vaginosis infections, soil, and tree endophyte samples to discover and assemble the genomes of novel strains of known species as well as novel prokaryotes and eukaryotes.  The high quality of Hi-C-based assemblies allows the simultaneous closing of numerous unculturable genomes, placement of plasmids within host genomes, and microbial strain deconvolution in a way not possible with other methods.