Our Research

Developing and applying data-driven computational approaches to discover and exploit novel mechanisms of biological signal transduction, using a combination of structural, evolutionary and network bioinformatics, as well as machine learning techniques.

Data-driven discovery of signaling mechanisms

A major focus is to learn principles of signaling mechanisms in systems such as G protein Coupled Receptors (GPCRs) in cancer or the Leucine-rich repeat kinase 2 (LRRK2) in Parkinson Disease. We analyze large set of experimental interactions networks through bioinformatics and machine learning techniques to learn the sequence and structural determinants of binding as well as to develop predictor of interaction.

Selected publications:

Omics data analysis and interpretation for personalized medicine

We leverage biological mechanism knowledge and rich information from omics datasets to develop and apply machine learning pipelines to forecast disease phenotypes. Application ranges from using information from structure and interaction network to functionally annotate disease variants, to domain agnostic machine learning framework to learn omics signatures associated with a given phenotype.

Selected publications:

Drug and protein design

We also interested in exploiting this mechanistic knowledge to find via artificial intelligence new avenues to modulate biological systems either pharmacologically, via small molecule virtual screening, using representations of biological activity at different scales, or through the computational design of synthetic biologics with theranostic potential (e.g. Designer Receptor exclusively activated by designer drugs).

Selected publications: