Analyzing Alzheimer's disease gene expression dataset using clustering and association rule mining

摘要

Biological data like Gene expression datasets are already complex and are hard to process manually. The larger such types of datasets become, harder it becomes to manually process such datasets and makes more sense to use data mining techniques can be applied to discover or identify interesting patterns in the data. This paper presents various data mining techniques for analyzing Alzheimer's disease Gene Expression Dataset using Clustering and Association Rule Mining. The DNA-microarrays method allows acquiring a lot of data on gene expression. Due to the environmental and experimental factor, the variability of the gene expression is wide and unpredictable. This huge amount of data must be processed in order to retrieve relevant medical information. To do so, numerous methods of clustering are performed. There are two main goals: classify the gene expression and provide tools to retrieve the information. These techniques include basic data mining, two types of clustering and it discusses the use of association rules mining for such data. Emphasis is made on the particular dataset used in this research: the neurofibrillary tangles dataset that contains gene expression data for normal neurons and 'sick' neurons for ten different patients suffering from a mid-stage Alzheimer's disease.