Amino acid mutations in human proteins are often associated with inherited predispositions to specific diseases. Yet most observed missense polymorphisms, those involving a single nucleotide change leading to a changed amino acid, have not been characterized in terms of their effects on protein structure and function. We hypothesize that many of the most deleterious missense mutations affect protein function in one of two ways: 1) by altering interaction of proteins with other molecules, including other proteins, DMA, and small ligands; or 2) by altering stability of the protein. Both of these mechanisms depend primarily on the location of the mutation and its physical properties: changes in protein interactions are usually caused by mutations in or very near to a binding site; changes in stability are usually caused by mutations of buried hydrophobic residues.
The aim of this proposal is to develop a computational system for predicting the functional effects of missense mutations through homology modeling of protein complexes. New functional data on 1000 random mutations in two dimeric enzyme systems will be obtained to train and test the model. The primary application of this computational system will be to genes associated with the development of cancer. Cancer is usually linked to a number of genetic changes, some inherited and others somatic. These include loss of DNA-damage repair, breakdown of cell-cycle checkpoints, and resistance to apoptosis. Each of these processes requires many protein interactions, often in large protein complexes. These interactions may be compromised by missense mutations that alter individual interactions between molecules or mutations that lower protein stability.
Showing the most recent 10 out of 12 publications
