Untitled

Projects

COMETA — Cocoa Metabolomics

COMETA (COcoa METAbolomics) was a six-year interdisciplinary research initiative between Jacobs University Bremen and Barry Callebaut, aimed at mapping the roughly 20,000 molecules that give cocoa its taste, aroma, and flavor — from raw bean to finished chocolate. I joined the project in 2016 as a PhD student, contributing to two areas of the research: kinetic modeling of the fermentation process, and statistical consulting on a related metabolomics study. Learn more about COMETA.

Bayesian Kinetic Modeling of Cocoa Fermentation

My core PhD research focused on developing kinetic models of cocoa bean fermentation using Bayesian methods coded in Stan to fit systems of ordinary differential equations to previously published fermentation trial data. This work allowed us to quantitatively describe how key biochemical compounds evolve during fermentation — a critical step in determining the final flavor profile of chocolate.

I presented a summary of this research at StanCon 2019 in Cambridge. You can view the accompanying notebook here.

Tools: R, Stan, Bayesian inference, ODE systems

Statistical Consulting: Carbohydrate Kinetics During Fermentation

Alongside my main research, I collaborated with experimentalists in the group as a statistical consultant. One such collaboration led to a co-authored publication on changes in low-molecular-weight carbohydrates during cocoa fermentation, for which I applied Generalized Linear Mixed Models (GLMM) to characterize the differing degradation kinetics of these compounds.

Read the full paper: Food Research International (2020)

Tools: R, GLMM, statistical consulting