Exploring the genetic heterogeneity of Alzheimer’s disease: Evidence for genetic subtypes

Abstract

Alzheimer's disease (AD) has a complex and multifactorial etiology. It is under significant genetic influence, thus identifying systematic variation in genetic risk represents a potentially useful approach to understanding diverse origins of the disease. Here, we explore genetic heterogeneity underlying AD using a multi-step approach. First, a principal component analysis was applied to AD-associated variants across 2,739 AD cases and 5,478 age and sex-matched controls from the UK Biobank. Three distinct clusters (termed "constellations") were found, each containing a mix of cases and controls. This structure only emerged when restricting analysis to AD-associated variants and therefore appears to be disease-relevant. Next, we applied a recently developed biclustering algorithm that searches for subsets of AD cases and variants that constitute distinct risk groups. We found two significant biclusters, each manifesting disease-specific genetic signatures that increase risk for AD. This pattern of clustering replicated in an independent dataset from the Alzheimer's Disease Neuroimaging Initiative (ADNI). These findings reveal a hierarchy of underlying AD genetic risk. At the first level, disease-relevant constellations may represent differential vulnerability of specific biological systems or pathways that contribute to disease development but are not sufficient to increase disease risk on their own and likely require additional risk factors. At the next level, biclusters may represent disease subtypes, or groups of AD cases with unique combinations of genetic variation that increase their risk of developing AD. More broadly, this study illustrates an approach that can be extended to investigations of genetic heterogeneity underlying other complex diseases.