Abstract

Single nucleotide polymorphisms (SNPs) are the most common form of sequence variation within the human genome and occur once every 1000 bases. There are now many databases which catalog these variations. It is expected by analysing the sequence variations found within a population, associations can be made with disease which will accelerate the identification of disease genes and lead to a better understanding of the disease process. I have used an automatic comparative modelling procedure to build models of the protein structure for over 7000 genes. I will describe an analysis of sequence variations which cause amino-acid sequence changes (non-synonymous SNPs) within the coding regions of these genes. The first set of data, derived from the OMIM database, are known to cause Mendelian inherited diseases. The second set, from dbSNP, currently have no annotated link to a disease. In total there are 24,000 nsSNPs. I have applied structure-based as well as sequence-based tools to predict which of these sequence changes are likely to either disrupt the structure of the protein or interfere with the function or interactions of the protein. Some of these predictions are able distinguish disease causing mutations from those mutations which are thought to be have a neutral affect. I will give examples of mutations in genes which are predicted to be deleterious and may have a role in common human diseases.