A method for the culture-free discovery and host affiliation of novel viruses from metagenomic samples

ABSTRACT In this application, we propose to develop a novel reagent kit and accompanying analytic software platform that employs Hi-C technology for a user-friendly method to assemble viral genomes from metagenomic samples and associate these viruses with their microbial hosts. This product enables culture-free discovery of phages and quantitative measurement of their host range directly in mixed microbial communities. Bacteriophages are viruses that infect bacteria; they shape microbial ecosystems through predation on hosts and through horizontal gene transfer. They are also an important vector in the transmission of anti-microbial resistance. Despite their potent impact on microbial biology, only a tiny proportion of phage genomes are represented in public databases in part because of the difficulty of isolating phage whose hosts are not culturable laboratory settings. This limitation in our ability to understand viral biology has important impacts on emerging biotechnology applications including fecal microbiota transplantation and phage therapy as well as wide-ranging effects on microbiological research. We propose to develop a novel computational approach combined with improved experimental methods to deconvolute viral genomes and perform host attribution from whole microbiome samples without culturing. At the core of this approach is utilization of proximity ligation methods (Hi-C) which physically associates DNA sequences present within intact microbes. This provides direct physical evidence of the contiguity of viral genome sequences and host affiliation information that is not achievable by any other method. The outcome of this direct-to-Phase II proposal would be a combination of kit and user-facing computational platform that would empower both the sophisticated next-generation-sequencing biologist and the sequencing novice to ask important questions about viruses in their microbial community of interest. Aim 1 focuses on computational methods improve the recovery of viral genomes from metagenomic data. Aim 2 develops methodologies to enhance viral Hi-C chemistry. These two aims work together synergistically enhance data quality. Finally, Aim 3 develops a computational platform for data analysis through a web portal where results could be visualized directly via web browser or raw data downloaded for further exploration. Upon completion we will have developed a first-in-class commercial platform to discover viral genomes and identify their hosts from complex microbial communities.