Computational approach to detect the culprit genes responsible for a Disease

Computational exploration of the identity of genes which may be responsible for a particular disease is a very exhaustive study nowadays. This uses the datasets available in different websites which store a tremendous amount of genetic information in the form of microarray data. In this paper, we have presented a reliable and effective approach for unveiling the smallest possible set of genes which is associated with even a quite poor prognosis disease like Alzheimer Disease. The dataset used here has been collected from the official website of NCBI. We have used rough set theory, random forest, and principal component analysis or their collective form for the said purpose. The maximum accuracy achievable here for the purpose of diagnosis is quite satisfactory. Further, we have verified our result from DAVID ontological website where it has been found that most of the genes extracted computationally are really associated with Alzheimer´s disease.