Study Of SNP Disease Models And Their Detection Performances

Single nucleotide polymorphisms (Single Nucleotide Polymorphism) refers to thepolymorphisms caused by the variation of a single nucleotide (A, T, C, G) of thegenomic DNA sequence in the human genome, which is the most widely distributedpolymorphism and abandant in genetic information. There are a large number of SNPloci that determines the personality differences between people and be greatlyassociatied with complex disease, So SNP study is of great significance, and now SNPstudy is widely considered an important step towards the application of the HumanGenome Project.In this paper, we first study the nature of the disease model to study the relationshipof disease model among the parameters MAF (Minor Allele Frequency), MES(Marginal Effect Size), Popular Prevalence and other parameters in the classicaldisaese model we choosed. We also analyzed the influence of the parameters to theassociation measure between the disease locus and class label of the sample, andcompared them between the disease model and noise model. In addition,Corresponding experiments are designed based on the study of the nature of thedisease model before to compare detection ability of the current mainstreamassociation analysis algorithms AntEpiSeeker, SNPRuler, Boost algorithm on thesimulation data under different disease models and different parameters. Meanwhile,we did the experiments to compare the computing time among the different algorithms.Finally, we give the results and analysis of the expriments on real SNP data ofAge-related macular degeneration(AMD) and lung cancer.