RNA viruses are diverse components of global ecosystems.
Bulk RNA shotgun sequencing – metatranscriptomics – has transformed our understanding of the virosphere, providing a uniquely powerful means to describe the viral composition of any sample, and helping to reveal how viruses move across the human-animal interface and eventually emerge as new infectious diseases. However, the metagenomic identification of RNA viruses has traditionally been limited to those with sequence similarity to known viruses, such that highly divergent viruses that comprise the “dark matter” of the virosphere remain challenging to detect. Herein, I will show how metatranscriptomics, combined with advances in artificial intelligence (AI) technology that can integrate primary sequence and structural information to accurately and efficiently detect viral sequences, is providing new insights into fundamental aspects of virus evolution, ecology and emergence.
I will use metatranscriptomics to identify the fundamental drivers of virus diversity and evolution at the scale of individual ecosystems, revealing the impact of host barriers to cross-species virus transmission. I will also show how a combination of metatranscriptomics and AI has led to the discovery of tens of thousands of novel RNA viruses, redefining our knowledge of the scale and composition of the virosphere. The RNA viruses newly identified were far more divergent than those described previously, and present in diverse ecological niches, including the air, hot springs and hydrothermal vents, varying dramatically in abundance between ecological types.