Abstract:Objective To compare the risk assessment models of type 2 diabetes mellitus (T2DM) by logistic regression,Artificial Neural Network and SVM models and to provide the basis for screening,health management and individualized prevention of T2DM in residents with high risk.Methods A cross sectional study was carried out in Fangshan district of Beijing.A total of 745 patients with T2DM and 1 964 controls were finally selected as the subjects of the present study,The standardized quesdonnaire was used to collect the data related to socio-demographics,lifestyle factors and history of chronic diseases.The standard physical examination and biochemical test were performed for all subjects.The matrix-assisted laser desorption/ionization time-of-flight mass spectroscopy was utilized to detect the genotypes at 27 SNP loci of 15 T2DM susceptibility genes reported by GWAS.The genetic risk of all subjects was assessed by non-weighted genetic risk score method.Logistic regression,Artificial Neural Network and SVM models were used to construct the risk assessment models of T2DM,which included the genetic and environmental risk factors.Results (1)Logistic regression model showed that the traditional risk factors score was OR=0.647 and 95％CI:0.623-0.670,the genetic risk factors score was OR=0.579 and 95％CI:0.554-0.604 and the score of traditional risk factors with genetic risk factors was OR=0.670 and 95％CI:0.647-0.693 in AUC of the model.(2) Artificial Neural Network model showed that the traditional risk factors score was OR=0.690 and 95％CI:0.668-0.713,the genetic risk factors score was OR=0.579 and 95％CI:0.554-0.604 and the score of traditional risk factors with genetic risk factors was OR=0.728 and 95％CI:0.706-0.749 in AUC of the model.(3) SVM model showed that the traditional risk factors score was OR=0.733 and 95％CI:0.663-0.803,the genetic risk factors score was OR=0.562 and 95％CI:0.484-0.640 and the score of traditional risk factors with genetic risk factors was OR=0.765 and 95％ CI:0.700-0.831 in AUC of the model.Conclusion As compared with the traditional risk factors score in models,the score of traditional risk factors with genetic risk factors in models shows the better prediction for T2DM risk assessment.As compared with logistic regression model,Artificial Neural Network or SVM models show the better prediction for T2DM risk assessment.

Application Study of BPANN on Genetic Variants in Peroxisome Proliferators Activated Receptor-Γ (Ppar-Γ) and Retinoid X Receptor-Α(rxr-α) Gene and Type 2 Diabetes Risk in a Chinese Han Population

Application of Back Propagation Artificial Neural Network on Genetic Variants in Adiponectin ADIPOQ, Peroxisome Proliferator-Activated Receptor-γ, and Retinoid X Receptor-α Genes and Type 2 Diabetes Risk in a Chinese Han Population

Application of the Back-Error Propagation Artificial Neural Network （BPANN） on Genetic Variants in the PPAR-γ and RXR-α Gene and Risk of Metabolic Syndrome in a Chinese Han Population

The Pro~(12) → Ala Substitution in PPARγ_2 Gene May Be a Protective Factor for the Development of Type 2(diabetes) in Chinese Han Population

Genetic Variants in Peroxisome Proliferator-Activated Receptor-γ and Retinoid X Receptor-α Gene and Type 2 Diabetes Risk: A Case-Control Study of a Chinese Han Population

Uncoupling Protein 2 and Peroxisome Proliferator-Activated Receptor γ Gene Polymorphisms in Association with Diabetes Susceptibility in Chinese Han Population with Variant Glucose Tolerance

Tagging Single Nucleotide Polymorphisms in the PPAR-γ and RXR-α Gene and Type 2 Diabetes Risk: a Case-Control Study of a Chinese Han Population

Modulation of the Association between the PEPD Variant and the Risk of Type 2 Diabetes by n-3 Fatty Acids in Chinese Hans

Association and Interaction Analysis of PPARGC1A and Serum Uric Acid on Type 2 Diabetes Mellitus in Chinese Han Population.

Uncoupling Protein 2 And Peroxisome Proliferator-Activated Receptor Gamma Gene Polymorphisms In Association With Diabetes Susceptibility In Chinese Han Population With Variant Glucose Tolerance

Genetic variation in the peroxisome proliferator-activated receptor (PPAR) and peroxisome proliferator-activated receptor gamma co-activator 1 (PGC1) gene families and type 2 diabetes.

A Population-Based Study of PPARγ2 Gene Polymorphism for Type 2 Diabetes

Genetic Variations In Dipeptidyl Peptidase 4 Gene Have A Synergetic Effect With Body Mass Index On The Risk Of Coronary Artery Disease In A Chinese Han Population With Type 2 Diabetes

The Association of Between the Polymorphisms of the Adiponectin Gene Rs2241766 and Rs1501299 and the Metabolic Traits of Type 2 Diabetes in Han Chinese

Peroxisome Proliferator-Activated Receptor Delta (Ppard) Genetic Variation And Type 2 Diabetes In Middle-Aged Chinese Women

Association Between Genetic Polymorphisms in Exon 2 and 6 of the PPAR-γ Gene and Susceptibility to T2DM in Han Population of Jiangsu

Variants in PPARD- GLP1R are related to diabetic kidney disease in Chinese Han patients with type 2 diabetes mellitus

The Relationship Between RXRα Genetic Variation and Susceptibility of Type 2 Diabetes

PDX1 and MC4R Genetic Polymorphisms Are Associated with Type 2 Diabetes Mellitus Risk in the Chinese Han Population.

Association study of the single nucleotide polymorphisms in adiponectin-associated genes with type 2 diabetes in Han Chinese

Comparing the risk assessment models of type 2 diabetes mellitus on the basis of environmental and genetic risk factors