The Application Of Ant Colony Optimization On Finding The Optimal Gene Order Of Alzheimer's Disease

Alzheimer disease (AD) is the most common form of dementia, which is a progressive neurodegenerative disorder characterized by global cognitive decline involving memory, orientation, judgment, and reasoning. In 2006, there were 26.6 million sufferers worldwide. Alzheimer's is predicted to affect 1 in 85 people globally by 2050.At present, there is no definitive evidence to support that any particular measure is effective in preventing AD. In order to treat Alzheimer's disease, the genes causing Alzheimer's disease must be found to understand the pathogenesis of Alzheimer's disease, so the identification of the genes associated with AD is one of key studies. Currently many approaches to identify the genes associated with Alzheimer's disease are through analyzing DNA microarray data and different expression of genes. And one of the analyses is to align the genes that have similar expression levels together. Conventional approaches only consider the results, but never consider the alignments in cluster. So, in this paper, the problem of finding the optimal gene order of Alzheimer's Disease(AD) is first formulated as Traveling Salesman Problem(TSP), and applied Ant Colony Optimization(ACO) to solve the TSP. The main research contents are listed as follow:(1). Alzheimer disease, the technology of gene-chip, current research situation to identify genes associated Alzheimer disease and the main methods of finding the optimal gene order are introduced. (2). Firstly, the basic Ant Colony Optimization is applied the finding the optimal gene order of Yeast genes, and three measurements are analyzed and discussed in the experiments; secondly, information entropy is introduced into ACO and is set as the stop criterion of ACO, which can estimate the number of iteration and reduce the redundant iteration, and the results show that the larger the data scale is, the better the algorithm performs; finally, the ants in ACO are divided into two groups, and the scale is reduced during the iterations, also, information entropy is set as the stop criterion, which can get almost the same quality of gene order while reducing the running time greatly.