Novel method for discovering viral genomes in human metagenomic data

Published: 2018-01-19


A computational strategy for finding novel viral genomes developed by a research team led by Erik Sonnhammer (Stockholm University/SciLifeLab), Björn Andersson (Karolinska Institutet/SciLifeLab), and Tobias Allander (Karolinska Institutet) was used to identify a number of protein families, lacking previously known homologs, in human samples such as serum, feces and cerebrospinal fluid.

The study, published in Scientific Reports, identified a completely novel bacteriophage present in the digestive tract using PCR-based cloning and sequencing starting from one of the discovered families. The different bioinformatic approach is complementary to traditional characterization of virome datasets, which relies on homology-based methodologies. It presents new opportunities for understanding the role of the microbiome in health and disease.