An Information and Analysis System for Coronary Artery Disease (CAD) Diagnostic Gene Microarray

摘要

Motivation: Efficient, accurate and automatic prediction of the likelihood of a person getting a genetically predisposed disease, such as coronary artery disease (CAD), according to a prototype diagnostic DNA microarray, enabling high risk individuals to take early precautionary measures. Results: We present a two-stage CAD information analysis system to assess an individual's risk of heart disease based on results of genotyping by mini-sequencing using microarrays. In the first stage, a Convert computer program is applied to obtain the genotype profile for a patient from raw genotyping intensity data. The second stage makes use of several neural network and statistical techniques, such as multi-layer perceptron (MLP), logistic regression (LR) and hierarchical mixtures of experts (HME) to determine the probability of the subject having CAD. A web-based CAD analysis system has also been developed to integrate all the information.