Frank B. Hu

This paper develops an optimization-based dietary recommendation approach to improve diet scores using simulated annealing and evaluates it on Diet-Microbiome Association study data.

This study introduces METRIC, a deep-learning method that uses gut microbiome data to correct random errors in self-reported dietary assessments, showing effectiveness in simulated and real-world datasets.

The paper identifies nutritional redundancy in dietary intake and shows its association with lower risks of cardiovascular disease and type 2 diabetes after adjusting for age.

genuMet identifies and reduces false metabolic features in untargeted metabolomics without pooled QC samples, achieving high true negative and true positive rates.