Mapping global functional diversity of soil microbes using metagenomics data

Soil microorganisms play a critical role in global carbon fluxes and shape local biogeochemical cycles through their vast functional diversity, yet it remains unclear how this diversity influences soil carbon fluxes at the global scale.  For example, unlike plants, which are almost uniformly autotrophic, microbial communities encompass a wide range of substrate use : however, current models lack a simplified, yet representative framework to capture this functional diversity, limiting our ability to accurately predict biogeochemical cycling in a changing climate.

To address this, we propose a trait-based microbial functional classification that leverages the growing availability of metagenomic data. Using the microTrait tool, we analyze trait information from a global database of 40,000 metagenome-assembled genomes (MAGs) to compare several clustering methods with multiple quality metrics, and define ecologically meaningful functional groups.

By backmapping MAGs to their original metagenome, we obtain relative abundance data, allowing us to examine how microbial community composition varies across environmental gradients of soil, climatic and biotic parameters. Our analysis reveals some associations between community structure and environmental parameters, suggesting that integrating microbial functional traits into soil models could improve biogeochemical predictions.