This approach, developed by scientists at Roche, promises to accelerate the discovery of mouse correlates of genetic risk factors for human disease.
Because the mouse genome is similar to that of humans, the mouse is the most commonly used experimental model for studying human disease, and the 'mouse to man' approach is widely used. Since analyses of mouse genetic models by traditional methods are very time-consuming and costly, this novel computational approach represents a major advance for this entire field of research.
The new technique enables researchers to identify a single causative genetic factor by correlating a pattern of observable physiological or pathological differences among selected strains of mice with a pattern of genomic variation.
Using conventional methods, pin-pointing a gene contributing to disease risk could take five scientists five years. With Roche's latest innovation, which has up to 1,000-fold greater precision than current methods, a single researcher may accomplish the task in a single afternoon.
The method takes advantage of the block-like patterns of genomic variation in selected mouse strains.
"Our hope is that this new computational approach will increase the utility of the vast amount of DNA sequence information available today and help researchers more fully leverage mouse models of human disease to identify genes contributing to disease risk and drug response," said Gary Peltz, head of genetics and genomics at Roche's R&D facilities in Palo Alto, US.
He added that the technique will help researchers understand the relationship between trait differences and variations in the mouse genome, which will make it easier to understand the impact of human genetic differences.
"As that happens, we should be able to translate genetic data more effectively and efficiently into the development of both novel diagnostic tools and new medicines to treat human diseases."
The paper, entitled "In Silico Genetics: Identification of a Novel Functional Element Regulating H2-Ea Gene Expression," reports that the new computational algorithm correctly identified the genetic basis for strain-specific differences in several biologically important traits, including differences in drug metabolism.
The examples presented in the paper demonstrate the ability of the methodology to identify causative genetic factors accurately for a wide range of trait data. The technique also has the potential to uncover currently unknown genetic factors contributing to a host of different diseases.
Roche scientists first published a computational method for mouse genome analysis in the 8 June, 2001 issue of Science. That method predicted regions of a mouse chromosome responsible for a trait difference. The predicted regions contained hundreds of genes and the results were assessed by relative (percentile ranking) statistical criteria. The new method offers the same analytic speed, but is much more exact, linking a single gene to a trait difference.
Crucially, this method eliminates the need for follow-up studies to mine large chromosomal regions, saving researchers from months to years of experimentation.
The pattern of genetic variation analysed by this new computational method was created by mining a database of common genetic markers, called single nucleotide polymorphisms (SNPs), covering 1,900 genes across 16 commonly used inbred mouse strains.
The genetic pattern maps are now available to the public for the first time as part of the Roche SNP database web site.
Roche Palo Alto is engaged in research with several leading universities and government institutions to make use of the new computational technique. The studies are directed toward better understanding the genetic causes of a range of human diseases and toward pharmacogenetic analysis of how various drugs that are used commonly to treat disease work in humans.