Multivariate regression between vitamin D use and gut microbiome of young infants

This project aimed to build a generalized linear model to evaluate association between vitamin D use, fecal metabolites, and microbiota, adjusting for covariables. The project used a cohort dataset of 575 infants, including questionnaire, NMR spectrum and DNA seq data.

Machine learning classifiers for childhood obesity trajectrories using infant fecal microbiota data

Knowing obesity risk earlier could help healthcare professionals manage children’s weight more effectively. A GLMM-based classifier was built to predict children obesity with an AUC of 0.83, specificity of 73% and sensitivity of 85%, based on 2500 infants data.

WEGAN: Web Application Streamlining Ecology Data Analysis

This collaborative project aims to develop a web-based software application that automates data analysis for ecologists and produces reproducible reports within minutes.