Numerics

I have always been mesmerized by how a computer could execute repetitive tasks for people. I have thousands of bits of code doing things here and there; most are written Python or C++. Still, the remarkable ones revolve around subjects down below.

Mechanistic modeling & CFD

I develop codes that can reproduce biological behaviors (cell growth, metabolites accumulation, …) and couple them with full-scale CFD computation of bioreactors. This way, it is possible to predict what would happen to each and every cell of the population and manipulate it.

In addition, I apply my skills in mechanistic modeling and CFD to many other topics, such as heat and mass transfer in porous/granular media.

Statistics & Machine learning

Biology requires assessing the statistical robustness of the conclusions we draw from the data we generate. That is why I acquired skills in descriptive and inferential statistics.

Machine learning was the obvious next step. Nowadays, biotechnology high throughput methods, such as flow cytometry and increasing HPC power, make many things possible. I apply machine learning algorithms to three types of problems:

• Clustering (unsupervised learning) - to identity a sub-population of cells

• Spectrophotometric quantification (supervised learning) - to lower analytic response time and better operate bioreactors

• Data analysis - to understand the underlying couplings within the data using non-linear models