Modeling short-term antidepressant responsiveness with artificial neural networks

Abstract

correspondence: Po see chen Department of Psychiatry, hospital and college of Medicine, national cheng Kung University, Tainan, Taiwan Tel + 886 6-235-3535 ext 5190 Fax + 886 6-275-9259 email chenps@mail.ncku.edu.tw Abstract: Due to the varying nature of patient response to different types and even dosages of the same antidepressant, doctors currently prescribe antidepressants on a trial and error basis. Therefore, it is highly desirable, both clinically and economically, to establish tools that distinguish responders from non-responders and to predict possible outcomes of the antidepressant treatments. The overall effectiveness of treatment using antidepressants may thus be optimized. Common genetic polymorphisms, such as single nucleotide polymorphisms (SNPs) can be used in clinical association studies to determine the contribution of genes to drug efficacy. In this work we developed a prediction model resulting from the analysis of clinical factors such as SNPs, age, baseline Hamilton Rating Scale for Depression (HAM-D) score, antidepressant groups, and gender of depression patients. We used it to predict the responsiveness of antidepressant treatment. By using candidate genes reported in the literature, we selected four SNPs that were strongly relevant to antidepressant efficacy. Our study population consisted of Taiwanese patients with major depression recruited from the National Cheng Kung University Hospital. The genotyping data was generated in the high-throughput genomics lab of Vita Genomics, Inc. With the wrapper-based feature selection approach, we employed multilayer feedforward neural network (MFNN) and logistic regression as a basis for comparisons. Our data revealed that the MFNN models were superior to the logistic regression model. The MFNN approach provides an efficient way to develop a tool for distinguishing responders from nonresponders prior to treatments. Our preliminary results showed that the MFNN algorithm is effective for deriving models for pharmacogenomics studies and for providing the link from clinical factors such as SNPs to the responsiveness of antidepressants in clinical association studies.