Camille Champion, (MAP5, Université Paris Cité)
Modeling microbial interactions as sparse and reproducible networks is a major challenge in microbial ecology. Direct interactions between the microbial species of a biome can help to understand the mechanisms through which microbial communities influence the system. Most state-of-the art methods reconstruct networks from abundance data using Gaussian Graphical Models, for which several statistically grounded and computationnally efficient inference approaches are available. However, the multiplicity of existing methods, when applied to the same dataset, generates very different networks. In this presentation, I will introduce OneNet, a consensus network inference method that combines seven methods based on stability selection. We demonstrated the precision of our methodology using synthetic data and exhibited a meaningful microbial guild when applied on gut microbiome data from liver-cirrothic patients.