GUILLAUME LAMOUREUX

We are developing methods for computational enzymology and applying them to enzymes containing metal atoms at their active sites (zinc, iron, etc.). Although about half of all known proteins contain at least one metal atom, no method is currently available to efficiently and reliably simulate chemical reactions in metalloenzymes. We are focusing on enzymes of medical and biotechnological interest, such as cytochrome P450s and zinc hydrolases... We are developing fast computer algorithms to predict protein-protein interactions and to infer which mutations are likely to perturb these interactions (and affect biological function). We are taking advantage of a number of recent developments in a field of machine learning called deep learning. Deep learning technologies are commonly used for automatic translation and for voice or image recognition. Treating the interaction of two proteins as a form of recognition - molecular recognition -, we are developing deep learning algorithms to..