Biomolecular modelling and design

Research topics

The research of the Biomolecular Modeling and Design group is dedicated to the development and application of computational protocols to evaluate and optimize the features of protein systems, with special focus on the design of protein binders, such as peptides and single-domain antibodies. We combine a plethora of theoretical approaches, such as Monte Carlo, docking, molecular dynamics, and machine learning, into a state-of-the-art mutagenesis evolutive platform for the optimization of binder-target affinity. In addition, we are interested in the computational prediction of other important physicochemical features of proteins. This area complements our ultimate goal of establishing a comprehensive computational suite for protein fitness prediction and design, thereby helping to overcome the most common difficulties faced by biomedical and biotechnological applications. Our research covers a wide range of disciplines, such as computational biochemistry, biophysics, bioinformatics, and physical chemistry. The group is currently interested in the following topics:

Peptide design

Nowadays, peptide binders are involved in a wide variety of biotechnological and pharmaceutical applications. This scenario calls for more efficient and versatile peptide design strategies, where computational tools offer interesting alternatives. Our computational protocol has proven successful in the design of cyclic peptides used for various experimental applications, such as molecular recognition probes, aggregation inhibitors, or immobilization agents.

Nanobody design

Single-domain antibodies, or nanobodies, are the smallest units capable of performing molecular recognition tasks similar to those of standard antibodies. Their small size is a significant advantage for their engineering using in silico protocols, which allow the design of binders capable of capturing their target through a chosen binding site. This goal is now at reach, as demonstrated by our results in the de novo design of antibody fragments specific for the protein epitope, already validated in wet lab.

Computational prediction of protein features

The variation of a single amino acid in a protein can completely alter its structural and functional characteristics, causing serious pathological disorders or a critical loss of stability. Predicting the impact of a protein mutation using computational tools is significantly advantageous, as it would avoid the need for time-consuming experimental validations. Our computational methods have proven effective in the molecular description of the effects of missense mutations in various neurological disorders, as well as in predicting the stability and expression yield of engineered protein binders.