We demonstrated that a trained artificial neural network model is a promising method for providing the inference from factors such as single nucleotide polymorphisms, viral genotype, viral load, age and gender to the responsiveness of interferon.

The relationship between obesity and a single nucleotide polymorphism (SNP), rs5443 (C825T), in the guanine nucleotide binding protein beta polypeptide 3 (GNB3) gene is currently inconsistent. In this study, we aimed to reassess whether the GNB3 rs5443 SNP could influence obesity and obesity-related metabolic traits in a Taiwanese population. A total of 983 Taiwanese subjects with general health examinations were genotyped. Based on the criteria defined by the Department of Health in Taiwan, the terms ''overweight'' and ''obesity'' are defined as 24 2 BMI \ 27 and BMI 3 27, respectively. Compared to the carrier of the combined CT ? TT genotypes of the GNB3 rs5443 polymorphism, triglyceride was significantly higher for the carrier of CC genotype in the complete sample population (128.2 ± 93.2 vs. 114.3 ± 79.1 mg/dl; P = 0.041). In addition, the carriers of CC variant had a higher total cholesterol than those with the combined CT ? TT variants (194.5 ± 36.8 vs. 187.9 ± 33.0 mg/dl; P = 0.019) in the complete sample population. In the normal controls, both triglyceride (P = 0.018) and total cholesterol (P = 0.011) were also significantly higher in the CC homozygotes than in the combined CT ? TT genotypes. However, the GNB3 rs5443 SNP did not exhibit any significant association with obesity or overweight among the subjects. Our study indicates that the CC genotype of the GNB3 rs5443 SNP may predict higher obesity-related metabolic traits such as triglyceride and total cholesterol in non-obese Taiwanese subjects (but not in obese subjects).

The SVM model is a promising method for inferring responsiveness to IFNalpha dealing with the complex nonlinear relationship between factors (such as SNPs and viral genotype) and successful therapy. In this study, we demonstrate that our tool may allow patients and doctors to make more informed decisions by analyzing host SNP and viral genotype information.

A genetic model was constructed to predict outcomes of the combination therapy in CHC patients with high sensitivity and specificity. Results also provide a possible process of selecting targets for predicting treatment outcomes and the basis for developing pharmacogenetic tests.

We demonstrated that our artificial neural network-based approach is a promising method to assess the gene-gene and gene-environment interactions for interferon and ribavirin combination treatment in chronic hepatitis C patients by using clinical factors such as SNPs, viral genotype, viral load, age and gender.