Mapping out the Alzheimer's disease data landscape

Abstract

A systematic in-depth comparison of multiple Alzheimer's disease (AD) clinical datasets was performed to evaluate how similar or distinct the datasets available to the AD research community are. Comparable datasets well suited for validation of hypothesis generated on the respective other were identified. Using a newly developed automated comparison framework "DataComp" feature overlaps, statistical significance of differences in feature distributions and biomarker progression across datasets were assessed and visualized.

A publication of the results is in planning.