Harnessing genomic data to predict functional traits for microbial communities

Microbial traits are difficult to measure, limiting the utility of trait-based biogeochemical models, especially at large spatial scales. Yet, we increasingly find that microbial communities control biogeochemical transformations, especially in soils where a majority of terrestrial carbon is stored. Microbial traits can be estimated from genomic data, and these traits are often phylogenetically clustered. Using algorithms to predict sparse trait datasets, we can harness existing measurements of microbial functional traits and their phylogenetic relationships to infer unknown traits. This method can be used to make inferences about the functional capacity of microbial communities under global change.