Finding novel viruses in lichen metagenomes

Marley 230

Abstract: Viruses infect all domains of life and are highly adapted to their host and their environmental niches. Despite the growing body of research on the role and impact of bacteriophage populations across various ecosystems, identifying viral sequences within large metagenomic datasets remains a significant challenge. The absence of a universal gene marker for viruses means that bioinformatics methods often depend on sequence homology searches against reference genome databases. Recent advances in bioinformatics have introduced machine learning-based tools designed to identify features indicative of phage origin, aiming to uncover previously unknown viral sequences. Here, we evaluate different machine-learning methods to investigate novel viral communities associated with lichens. Lichens form a highly diverse group, with more than 20,000 different known species and are able to colonize a vast range of habitats. These complex composite organisms arise from the mutualistic relationship of an algae or cyanobacteria living among filaments of multiple fungi. In this context, the potential role of viruses is currently unknown and largely unexplored. We apply cutting edge bioinformatic tools to detect viruses in a collection of lichen metagenomes, to investigate the diversity of the viral population associated with lichens and their role in the complex mutualistic association forming lichens.