I use computational approaches to study the behavior of biomolecules and their interactions, including structure, dynamics, kinetics, and thermodynamics. The goal is to understand and quantify the molecular mechanisms that govern the function of proteins and their complexes in normal or pathological conditions, particularly infectious diseases. See publications on the group's webpage.

I. Protein Interfaces and Molecular Recognition: The cooperative behavior of biomolecules is central to the function of all living organisms. Perturbations to the balance of intermolecular forces may lead to disease or, in the case of pathogens, to arrested growth or cell death. I am interested in protein interfaces, as they control almost all biologically relevant processes, including molecular recognition, association/dissociation mechanisms, enzymatic reactions, and the structure and dynamics of complexes. Most antimicrobials (be it against bacteria, viruses, fungi, or parasites) target one or more of these processes. My current work focuses on developing methods and algorithms to systematically search for critical steps in biological networks that could be perturbed to affect the network's behaviors deleteriously. This approach is a mechanism-based alternative to traditional high-throughput virtual screening and will enable the identification of novel protein targets and, potentially, new drug classes. Accurate modeling of protein interfaces, including the effects of the aqueous solution, is thus essential for a meaningful quantitative description.

II. Protein-Protein Interactions and Multiprotein Complexation: Mechanistic insight into protein-protein interactions, macromolecular organization, or network behavior in biological environments requires dealing with multiprotein systems for which reliable computational approaches are lacking or inefficient. Multiscaling (adaptive multiscale) simulations can address these problems efficiently. Such techniques are needed if the goal is to understand how interfering with specific protein-protein interactions might affect the behavior of a larger complex or process of which they are a part. Indeed, the topology of a network can be modified by targeting specific vertices, edges, or hubs through the action of one or more compounds. If the network attack is optimized and sustained, it might, depending on the network's robustness, shut it down entirely or severely impair it, both desirable strategies against pathogens. Suitable compounds include small molecules (traditional drugs) and macromolecules, such as antibodies, and emerging agents, such as nanobodies, cyclic peptides, and nanoparticles. Various computational approaches can be used for inverse design, including machine learning- and forcefield-based methods; my current interest is the design of branched, polycyclic compounds targetting shallow protein surfaces and cell membranes.

III. Methods Development: I use various theoretical and computational techniques, including molecular and mathematical modeling and simulations. A central aspect of my work is developing efficient models, methods, and algorithms to tackle problems at increasing levels of molecular complexity. The studies described in I and II often require new computational techniques because molecular events in biology involve time and size scales that span several orders of magnitude, from the purely atomistic to the thermodynamic and hydrodynamic regimes. In addition, since experiments are difficult to perform and molecular interpretations are often elusive, I conduct simulations to understand the underlying mechanisms and use the insight gained from these 'experiments' as guidelines for developing physics-based models and algorithms. I use well-established methods, including Molecular Dynamics and Monte Carlo, and ad hoc adaptations often motivated by the need to address specific experimental situations or simulation challenges.