Clustering Alzheimer’s Disease Gene Expression Dataset Reveals Underlying Sexually Dimorphic and Disease Status Profiles

Abstract

Background: The multiple appearance phenotypes in Alzheimer’s disease (AD) are manifested in epidemiologic sexual dimorphism, variation in age of onset, progress, and severity of the disease. Objective: In this study, we focused on sexual dimorphism, aiming to untie some of the complex interconnections in AD between sex, disease status, and gene expression profiles. Two strategic decisions guided our study: 1) to value transcriptomic multi-layered profiles over alterations in single genes expression; and 2) to embrace a sexual dimorphism centered approach, as we suspect that transcriptomic profiles may dramatically differ not only between healthy and sick individuals but between men and women as well. Methods: Microarray dataset GSE15222, fulfilling our strict criteria, was retrieved from the GEO repository. We performed cluster analysis for each sex separately, comparing the proportion of healthy and AD individuals in each cluster. Results: We were able to identify a biased, female, AD-typified cluster. Furthermore, we showed that this female AD-typified cluster is highly similar to one of the male clusters. While the female cluster constitutes mostly sick individuals, the male cluster constitutes healthy and sick individuals in almost identical proportion. Conclusion: Our results clearly indicate that similar transcriptomic profiles in the two sexes are “physiologically translated” in to a very different, dramatic outcome. Thus, our results suggest the need for a sex-based and transcriptomic profile-based study, for a better understanding of the onset and progression of AD.