A network-based integrated framework for predicting virus–prokaryote interactions

• DOI: 10.1093/nargab/lqaa044
• Publication Date: 2020-06-23T14:06:36Z
• Authors: Weili Wang; Jie Ren; Kujin Tang; Emily Dart; Julio Cesar Ignac...; Jed A Fuhrman; Jonathan Braun; Fengzhu Sun; Nathan A Ahlgren
• Keywords: Prokaryote

Metagenomic sequencing has greatly enhanced the discovery of viral genomic sequences; however, it remains challenging to identify host(s) these new viruses. We developed VirHostMatcher-Net, a flexible, network-based, Markov random field framework for predicting virus-prokaryote interactions using multiple, integrated features: CRISPR sequences and alignment-free similarity measures ([Formula: see text] WIsH). Evaluation this method on benchmark set 1462 known pairs yielded host prediction accuracy 59% 86% at genus phylum levels, representing 16-27% 6-10% improvement, respectively, over previous single-feature approaches. applied our tool crAssphage, human gut phage, two metagenomic virus datasets: marine viruses contigs recovered from globally distributed, diverse habitats. Host predictions were frequently consistent with those studies, but more importantly, made many confident than tools, up nearly 3-fold (n > 27 000), expanding diversity virus-host interactions.