Paper Title
Determination of The Risk SNP Associations in Akt1 and Met Genes From Osteoporosis

Abstract
The single nucleotide polymorphisms (SNPs) of AKT1 and MET genes have been reported to associate with osteoporosis. However,most of the studiesregarding to the association between loci remain to be determined. In this study, based on a top five results conservation strategy for multi-loci genotypes,we proposedan improvedgenetic algorithm to detect the high-order SNP associationsbetweenthe cases and controls data of osteoporosis. Five important SNPs amongst two genes (AKT1 and MET) in high free erythrocytes protoporphyrin (FEP) and low FEP subjects were determined and the test SNPs showed non-significant difference between the two subjects within single loci analysis. However, in 2- to 5-loci models of SNP associations, the SNPs of AKT1and MET genes are associated with osteoporosis (odds ratio [OR] = 2.119~3.138; 95% confidence interval (CI) 1.077~10.253; P-value< 0.005). The resultsrevealed that the SNP associations in osteoporosis can bedetermined by four SNP association models. AKT1 and MET genes possess the overall effects with susceptibility to osteoporosis;especially, the SNPs of rs1130233 andrs41735play animportant effect associated with osteoporosis.In conclusion, our proposed algorithm can successfully determine the SNP associations between genes. Key words: Osteoporosis, Single nucleotide polymorphism, SNP associations; Genetic algorithm.