Alzheimer's disease heterogeneity explained by polygenic risk scores derived from brain transcriptomic profiles

Area of Science:

• Neuroscience
• Genetics
• Computational Biology

Background:

• Alzheimer's disease (AD) presents significant clinical and neuropathological heterogeneity.
• Understanding the genetic underpinnings of this heterogeneity is crucial for developing targeted therapies.

Purpose of the Study:

• To investigate if polygenic risk scores (PRSs) combined with brain transcriptome data can explain AD clinical heterogeneity.
• To identify specific gene modules associated with AD subtypes and neuropathology.

Main Methods:

• Co-expression network analysis was performed on AD brain transcriptome datasets.
• Module-based PRSs (mbPRSs) were computed for gene sets associated with neuritic plaques and neurofibrillary tangles.
• Associations between mbPRSs and cognitive performance, clinical AD subgroups, and brain imaging data were tested.

Main Results:

• Two gene modules (M6 and M9) showed significant associations with AD neuropathology.
• mbPRSs from these modules correlated with distinct cognitive domains (language and visuospatial functioning), respectively.
• These findings were supported by regional brain atrophy data, aligning with specific clinical subtypes.

Conclusions:

• Polygenic profiling using co-expressed gene sets can effectively explain heterogeneity in Alzheimer's disease patients.
• This approach enables genetically informed patient stratification.
• The findings support the development of precision medicine strategies for AD.