Collaborative Project: ABI Innovation: Computational Identification & Screening for Deleterious Mutants

Iowa State University, Ames IA

Investigators

Abstract

When interpreting the effect of changes in genome sequences on their protein products, what are the criteria for separating the beneficial from the neutral and destructive mutations? It is not practical to experimentally test all possible changes; it is possible to develop computational rules and methods to accurately predict the effect of most changes, reserving experimental testing for borderline cases. The rules that constrain functional protein sequences, the structures into which they fold, the stability of the structures and the transitions that occur can be organized into an informatics framework that helps distinguish functional from non-functional mutations. The stability and dynamic transitions of protein structures over many mutations are the focus of this research, for which new evaluation methods will be developed and tested. Once the assessment methodologies are validated they will be integrated, along with the raw data and assessment results, within a single, efficient computational platform. The ability to interpret new data and compare it to existing results will not only allow discrimination of beneficial, neutral and deleterious mutations, but provides a resource for a protein-sequence-structure-based understanding of evolution. Informed selection rules will enable better decisions about the importance of saving individual members of endangered species, as well as how different environments affect selection in individual species. There is a large body of data in protein sequences and structures that can aid in understanding evolution, which is not currently being used. This project will utilize this data in systematic ways to shed new light on evolution. Protein structures are often modeled in different computational ways; the evaluation of these structure models is critical for expanding the set of reliable structures, since there are huge numbers of possible mutant sequences. Changes to the protein sequences and structures are not fully understood. Part of the difficulty lies in understanding the interactions within their densely packed structures, which have significant interdependences. Developing new ways to evaluate the effects of dense packing on protein structures is one of the project?s aims. Focusing on the physically interacting clusters within protein structures together with their sequence variants provides rich information about the correlations among amino acid substitutions. This rich data will then significantly advance the ability to distinguish between the important and the unimportant mutations. Progress has been seen with these approaches in the evaluations of predicted structure models at the CASP competitions for protein structure prediction. These new approaches lend themselves directly to the evaluation of the stabilities of different protein mutants. The evaluation procedures are derived directly from the available sets of protein structures and carefully tested against other known structures. In one important innovation these now include entropies that account for known changes in the structures of individual proteins, i.e. their dynamics. Capturing these tendencies for changes significantly improves the evaluation of the stabilities of proteins. Applications to sets of mutants show that unfavorable mutants are either more stable or less stable than the normal cases. These changes in stability directly affect the ways in which the proteins can move to carry out their functions; evaluating these changes significantly aids the understanding of protein mutations. This project will yield a uniform way to reliably assess the effects of protein mutations. This ability will significantly aid in the understanding of many aspects of evolution that remain.