Inferring Pathogenicity to Prioritize Candidate Disease–Causing Sequence Variations

Abstract

Abstract: : Background: Discovering the genetic causes of disease often requires one to search the genomes of a large number of patients and controls for variations that have a detectable impact on the organism's phenotype. The phenotype–altering variations that have been recognized to date are not uniformly distributed in the genome, in genes, or in the population. Thus, in most experiments designed to find these variations, the consumption of resources is non–linearly related to the likelihood of finding a phenotypically meaningful variation. Purpose: The purpose is to determine whether our emerging knowledge of conserved motifs, functional domains and SNPs could be used to predict regions of genes that would be more likely to harbor phenotype–altering variations. Methods: We developed a bioinformatic technique for prioritizing regions of genes for screening. The Prioritization of Annotated Regions (PAR) technique utilizes: 1) conserved protein functional domains, 2) protein secondary structures, and 3) SNPs, to predict the likelihood that a specific coding region of a gene will harbor a phenotype altering variant. Results: The PAR strategy was applied to 710 genes for which 4,498 previously identified mutations were known. Using regions of genes of size 200 nucleotides (approximating the optimal product size of an SSCP PCR amplification product size, excluding primers), 819 mutations were identified in 350 genes. This is 18% of the mutations correctly identified in approximately ½ of the transcripts. More importantly, of the 1,908,911 nucleotides represented by the 710 genes, the PAR strategy prioritized and selected 168,980 nucleotides, which represents a 91% reduction of screening resources compared to a comprehensive screening approach. Conclusions: These results suggest that prioritization strategies such as PAR can accelerate the success of mutation identification by more efficiently utilizing screening resources.

Keywords: gene screening • mutations • candidate gene analysis