Optimizing Viral Discovery in Bats

Viral discovery studies in bats have increased dramatically over the past decade, yet a rigorous synthesis of the published data is lacking. We extract and analyze data from 93 studies published between 2007–2013 to examine factors that increase success of viral discovery in bats, and specific trends and patterns of infection across host taxa and viral families. Over the study period, 248 novel viruses from 24 viral families have been described. Using generalized linear models, at a study level we show the number of host species and viral families tested best explained number of viruses detected. We demonstrate that prevalence varies significantly across viral family, specimen type, and host taxonomy, and calculate mean PCR prevalence by viral family and specimen type across all studies. Using a logistic model, we additionally identify factors most likely to increase viral detection at an individual level for the entire dataset and by viral families with sufficient sample sizes. Our analysis highlights major taxonomic gaps in recent bat viral discovery efforts and identifies ways to improve future viral pathogen detection through the design of more efficient and targeted sample collection and screening approaches.