Multi-factor SNP Association Algorithi Based On Stability

Because of its wide distribution, numerous amount and high mutation rate, SingleNucleotide Polymorphism (SNP) is worth to be studied in the fields of pathogenic genelocation and pharmacogenomics. SNP Association Study locates the locus whichinfluences disease from the perspective of the association between SNP loci sets withspecific disease.The situation of multiple pathogenic factors in the SNP Association Study isanalyzed in this paper. The similarities and differences between the SNP AssociationStudy and the Multi-factors SNP Association Study are also indicated. In order toimprove deficiencies of traditional SNP Association methods, a multi-factor discoverymethod which is based on the measure of the association’s stability is proposed. Aftergiving a comparison of some traditional SNP Association methods, a improved form ofmulti-factor discovery method which is based on the AntEpiSeeker algorithm isproposed. Deficiencies of traditional SNP Association methods are verified viasimulated data experiments. In addition, the performance of AntEpiSeeker, SNPRuler,SNP multi-factors discovery and SNP multi-factors discovery based on AES isevaluated via experiments. Finally, we apply our methods to an Age-related MacularDegeneration(AMD) and lung cancer data sets which are real SNP data sets, and we didfind some potential pathogenic factors with high reliability.