Polygenic Risk Scores are Associated to Aβ-τ Protein Accumulation in Alzheimer's Disease

Alzheimer's (AD) is a complex and multi-domain neurodegenerative disease characterized by amyloid-β (Aβ) and proteins aggregation holding both genotype and phenotype signatures. Increasing evidence suggests that the accumulation of -τ proteins plays a crucial role in disease onset and progression. In order to shade light on these issues, in this study we exploited the Partial Least Squares model to investigate the link between neuroimaging-derived features and genetic factors. Cortical thickness and Polygenic Risk Scores were used as pheno/geno-typing features, respectively. The model was trained on a cohort of subjects including healthy controls (CN) and AD patients, and tested on a discovery set consisting of - τ - and <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup>τ <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup>individuals. The working hypothesis being that these last two groups with different biological profiles would cluster closer to the CN and AD classes, respectively. In agreement with previous results, segregation was observed in the latent space between CN and AD and was also confirmed when applying the model to the - - and <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup> <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">+</sup>discovery set. These outcomes, together with a significant association found among the respective projection weights, encourage further investigation.