Jacob Bien

The paper develops a maximum-likelihood framework to estimate gene prevalence from incomplete genomes and applies it to Prochlorococcus, revealing a pangenome dominated by rare genes.

The paper introduces interaction models for microbial data to predict community functions and resistance genes, showing improved performance in diverse datasets.

The paper introduces a tree-guided aggregation framework for microbiome data to improve predictive modeling by learning data-adaptive taxon groupings.