SNP Data Simulation And Comparison Of Association Analysis Algorithms

Single nucleotide polymorphism (SNP), one of the most common types of DNApolymorphism, is referred to genome single nucleotide variation. Because it changed theoriginalstructure and chain rateofgene, so the sick probability ofthe individual increases.At present it has many association analysis algorithms which are about the analysis of therelationship between the SNP and the disease, but all these association analysisalgorithms lack of convincing due to lack of lots of SNP datasets with prior knowledge toprove their effectiveness.This paper studies the features of the true SNP data. After that, we establish a SNPdata simulation systemand produce some experiment datasets using this system, we havedone many experiments to prove our simulation SNP datasets to be effective, includingmutual information verification, multilayer dimensional reduction association analysisalgorithm verification and haploview software verification. Except these, this paper alsoconcentrates on comparison between several classical association analysis algorithms,such as BOOST association analysis algorithm, AntEpiSeeker association analysisalgorithm, SNPRuler association analysis algorithm, all these association analysisalgorithms have sophisticated softwareonthe internet, weproduce sevenSNPdatasetsbyour simulation system, and do experiments with these association analysis algorithms onthese simulation datasets.At last, wewillgive the influence ofthree important parametersof our simulation data and performance of these algorithms.