Detecting viral sequences in NGS data

Next generation sequencing (NGS) technologies provide an increasingly important avenue for detecting known viruses, and for discovering novel viruses present in clinical or environmental samples. Several computational pipelines capable of identifying and classifying viral sequences in NGS data have been developed and used to search for viruses in human or animal samples, microbiomes, and in various environments. In this review we summarize the different approaches used to determine viral presence in sequence data. Strategies for avoiding confounding factors such as physical contamination and computational artifacts that lead to false virus identification are discussed. The application of these methodologies to cancer data sets has led to important insights on viruses both as drivers of and biomarkers for specific tumors.