We develop computational methods for the analysis of proteins and use them to study evolution and predict protein spatial structures and functions. Our major recent advances are: 1) Statistically sound similarity search approaches based on matching of two sequence alignments augmented with known relationships between proteins in a database (COMPADRE); 2) Multiple sequence alignment program that is accurate for very distant sequences (PROMALS3D); 3) Comprehensive evolutionary classification of protein domains with known spatial structures (ECOD), a database that is updated weekly and catalogues most distant evolutionary connections between proteins; 4) Highly heterozygous genome sequencing, assembly and annotation pipeline that resulted in several dozen butterfly genomes sequenced by our group; 5) service to scientific community by being assessors in several structure prediction (CASP) and genome interpretation (CAGI) challenges. During the next 5 years, we would like to capitalize on our progress in the analysis of proteins and organisms and explore new research directions offered by developing technologies. Our work will be structured along the five major interconnected threads. 1) We will continue developing comprehensive evolutionary classification of proteins. We have the strongest track record and most extensive experience in this direction and are uniquely positioned to make a lasting impact. 2) We will develop computational methods to find distant protein homologs and multiply align them. This work will build upon our software we have been working on for almost two decades. These new methods will be used for protein classification, and expert analysis of protein families will offer fresh ideas for further methods development. 3) We will build the atlas of human mutations and rationalize their effects on proteins and human health. This project will rely upon our expertise in structure prediction, alignment and evolutionary connections between proteins, and will derive power from our dozens of collaborators, many of whom are clinicians who are dealing with interpretation of mutation effects. 4) At the organismal level, we will tackle the link between genotype and phenotype and a series of evolutionary and population biology questions using butterflies as model organisms. Many of these features are linked to proteins and their spatial structures. Integration of molecular biophysics techniques with organismal and evolutionary biology is innovative is promising to advance both fields. 5) We will continue collaborations with experimentalists to test our method and help them with experimental design. All five threads are tied by their connection to computational analysis of proteins, which is the main strength of our group.
Predictive computational methods are essential for understanding of biological phenomena and for experimental design. Our research program combines methods development with applications to the analysis of protein molecules and genomes. Such predictive analyses shed light on the molecular mechanisms of diseases and genetic disorders.
| Medvedev, Kirill E; Kinch, Lisa N; Grishin, Nick V (2018) Functional and evolutionary analysis of viral proteins containing a Rossmann-like fold. Protein Sci 27:1450-1463 |
