Home page - Eran Eyal

My main interests are protein dynamics, detection of correlation between residues based on evolutionary records (correlated mutations) and creation of scoring functions and automatic procedures for protein modeling, predicting consequences of amino acid substitutions and thermostability.

Research interests

Covariance analysis in proteins based on evolutionary records. Multiple sequence alignments (MSAs) are a reach source of information for protein analysis. I have developed a method to detect association between positions in the alignment (P2P). The method can help to detect sites in physical contact and help to discriminate between correct and in correct models. I used this and other available methods for detect covariance for analysis of HIV-1 protease, Glutamate transporters and other proteins.

Protein Dynamics. Elastic Network Models are simple yet powerful tool to explore equilibrium dynamics of macromolecules. It turned out they can assist in understanding large scale conformation changes, interactions and mechanical resistance of biomolecules. I have developed a user friendly interface for calculation and visualization of the equilibrium motions, extensively evaluated and refined the models based on X-ray data and applied the models to investigate mechanical properties of molecules.

Side Chain Modeling. Side chain modelling is required in a variety of tasks in structural bioinformatics and computational chemistry. For example, it is the final stage for construction of protein models by all major approaches for this problem: homology modelling, fold recognition and ab initio methods. It is also required for structural modelling of point mutations, predicting stability changes of proteins upon mutations and to complete missing crystalographic data regarding side chain position. Side chain modelling is also incorporated within programs for ligand docking. Our method based on contact surface areas is among the most accurate in the field and relatively efficient. It is publicly available as both stand alone application and a web server. Recently we used the method to model and evaluate RUN domain proteins.

Prediction of Stability Changes. Prediction of mutant stability is needed to understand the consequences of mutations and in order to design stable mutants. We calibrated and implemented our atom level surface-area-based scoring function for this task. In addition test other scoring functions, working on the residue level, to the same end.

Crystal Packing Analysis. Consideration of crystal contacts is important in order to examine protein-protein interfaces and in order to gain full understanding of protein observed conformations, especially on the protein surface. We developed a web server for analysis of inter molecular crystal contacts and surveyed the influences of crystal packing on protein structure and the associated hetero-groups.

My web Sites