Alzheimer's is a growing problem within today's society. The risk of Alzheimer increases exponentially with each decade an individual ages, and its fatality rate is hundred percent. Furthermore, no concrete methods of identifying the development of Alzheimer during a patient's lifetime exists, only through post-mortem analysis can we definitely conclude if the victim succumbed to Alzheimer's. Additionally, if the symptoms of Alzheimer's can just be offset by merely five years, trillions of dollars can be saved from healthcare costs (O'Connor 2010). The disease itself is fairly well documented: it has been shown that Amyloid- β proteins as well as phosphorylated Tau proteins are important in the role of developing Alzheimer's disease (Ihara 1986; Soscia 2010). Recent research provides that the buildup of amyloid- β plaques may be a result of an overactive immune system (Kolata 2010), and that the amyloid- β plaques themselves may even be anti-microbial particles that build up after being over-exposed (Soscia 2010). However, despite the research that has been done on Alzheimer's, little funding is going into researching a cure, or a way to lessen the impact of Alzheimer's on our aging population. Given the fact that every penny the National Institutes of Health spends on Alzheimer research, healthcare providers spend $3.50 caring for Alzheimer's patients, further coupled with the fact that every second, a baby-boomer reaches his 65th birthday, Alzheimer's is a problem that must be addressed immediately (O'Connor 2010). In order to address the problem, we have used Bayesian networks to identify which genes play an important role in Alzheimer's disease, and to illustrate the interactions among relevant genes. We have identified six human gene expression case control studies using microarray data and found 11 relevant genes that are related to Alzheimer's disease, four of which are strongly associated with Alzheimer's: TGFB1I1, LTF, TLX2, and LTB4R. We then analyzed this data, to identify pertinent interactions between the 11 genes. Using the data collected, we further calculated the lifetime risk, as well as the odd ratio of developing Alzheimer's disease, given the expression levels of combinations of genes. Females (aged 65+) with overexpressed TGFB1I1 and LTF, combined with under-expressed TLX2 and LTB4R showed a heighted 61% remaining lifetime risk of developing Alzheimer's disease, while males (aged 65+) with the same expression patterns of the genes showed a heighted 44% remaining lifetime risk of developing Alzheimer's disease. Furthermore, odd of developing Alzheimer's disease for an individual with normally expressed LTF and TLX2, under-expressed TGFB1I1, and over-expressed LTB4R is 25 fold higher than that of an individual with the same expression patterns of the genes except for an over-expressed TLX2. Given these findings we can help create effective methods of diagnosing patients with Alzheimer's disease and ultimately help create effective treatments for Alzheimer's patients.
