Exploring And Realization Of SNP Selection Strategy Based On DrSNP

With the completion of Human Genome Project(HGP), it's becoming possible for researchers to investigate the difference among individuals in genome scale. The most abundant source of genetic variation in the human genomes is single nucleotide polymorphisms(SNPs). Some of these variations may be the cause of the differences not only in susceptibility to disease but also in how people respond to medicine. So identification of the SNPs associated with Human disease phenotypes has great potential for direct clinical application by provding new and more accurate genetic markers for diagnostic and novel therapeutic targets. Under this background, our lab developed a microarray platform for parallel detection of one or more SNPs or mutation markers in a number of genomic DNA samples. As a part of this platform, in this paper, we construct a DrSNP( Disease-related SNPs ) database system and investigate the SNP selection strategies.During the research of SNPs associated with diseases, there is a large amount of data to deal with, including SNP information, experiments information and samples'information. In order to effectively store, search and analysis these data, we, in this article, design a DrSNP system. It contains three databases for storing the data of SNP, sample and experiment differently. With the help of this system, researchers can select proper SNP site and sample collection for experiment as well as obtain the statistical result of microarray experiments automatically.To select a few proper candidate SNP sites from over 9,000,000 human SNP sites for a microarray experiment is extremely important for all the SNP researchers. From the genome structure, we provide several selection methods. After that, we rank these methods with the difficulty of their realizations. These ranking system help us to build up a normal SNP selection strategy.Our selection strategy is based on the integration of genome sequences data and their annotation information. We download data from different international sources, such as NCBI Genebank, UCSC genome, dbSNP, dbTSS, peseudogene and Transfac. The information we need to conduct SNP selection is transformed from those primary data with different bioinformaticsmethods. In addition, we write some perl scripts to update the SNP date automatically.Using JSP technology, we realize the selection strategy on web. The selection page is designed to show both gene model figure and SNP information tables together. It enables us to select SNPs from gene, gene family or whole chromosome level. Meanwhile, it can serve as a framework for the future realization of other selection strategies.