Back

Identify Alzheimer's disease subtypes and makers from multi-omic data of human brain and blood with a subspace merging algorithm

Song, Z.; Huang, X.; Jannu, A. J.; Johnson, T. S.; Zhang, J.; Huang, K.

2025-05-07 bioinformatics
10.1101/2025.04.30.651565 bioRxiv
Show abstract

Identifying Alzheimers disease (AD) subtypes is essential for AD diagnosis and treatment. We integrated multiomics data from brain tissues of the ROSMAP and MSBB studies using a subspace merging algorithm and identified two AD patient clusters with notable cognitive and AD pathology differences. Analysis of differentially expressed genes (DEGs) in brain and blood samples pinpointed the LDLR gene as a potential blood biomarker linked to brain gene expression changes. Furthermore, we conducted PheWAS analysis on All of Us Projects EHR and WGS dataset for 105 eQTLs associated with the DEGs and revealed significant associations between these eQTLs and several phenotypes, shedding light on potential regulatory roles of these genes in diverse physiological processes. Our study successfully integrated multiomics data and proposes LDLR as a candidate blood biomarker for AD subtyping. The identified phenotypic signatures provide valuable insights on molecular mechanisms underlying AD heterogeneity, paving the way for personalized AD treatment.

Matching journals

The top 5 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.