Diversity in genes responsible for lifestyle-related diseases in Asia-Pacific region.

In Asia-Pacific countries, both environmental modernization and hereditary traits of Mongoloid reported to cause rapid increase in lifestyle-related diseases (LRD). However, reproducibility of reported responsive-factors is low. To examine this, a decision-tree method of complexity-model was applied to select LRD-responsive-factors. Genomic DNA was collected from Asia-Pacific regions. Single nucleotide polymorphisms (SNPs) on genomic DNA were determined as hereditary-trait-factor. Three indices of LRD (BMI, body fat, and serum leptin levels) were classified according to published criteria. WEKA Machine-learning system was used as decision-tree software. Age was added as a factor with different dimension. Selected factors were validated by other statistical methods. In Thai-males, GLUT) (glucose-transporter 1)-SNP was most-responsive to body fat, followed by USF1-SNP (transcription-factor for lipid metabolism). Differences between genotypes were validated (P = .002 for GLUT1 by Levene's, P = .071 for USF1 by ANOVA). Responsive-factors of Thai-females, Palau-males and Palau-females, were consisted with SNPs and age, and varied by groups. Convincing responsive-factors were not selected from mixed-data. Decision-tree-analysis successfully selected the convincing results. Responsive-factors differed by ethnic group and gender.