Multiomics machine learning identifies sleep and inflammation molecular pathways in prodromal Alzheimer′s Disease

Mild Cognitive Impairment (MCI) is a phase that can precede Alzheimer's Disease (AD). To better understand the molecular mechanisms underlying conversion from MCI to AD, we proposed a multiomics machine learning pipeline (four algorithms) to identify key pathways. Data consisted of metabolites (n=540) and proteins (n=3630) measured in blood plasma coupled with standard clinical tests (n=26). The cohort comprised 230 controls, 386 MCI participants and 184 AD-type dementia participants. Multiclass models showed that oleamide, MMSE and the priority language Z-score were the most relevant variables. Oleamide was increased in the MCI group and further increased in converters (both P<0.0001). In-vitro disease-associated microglia were able to synthesize oleamide and excrete it in vesicles. MCI conversion models showed pTau, tTau and JPH3, CFP, synuclein and PI15 proteins as the most relevant. This study uncovered molecular pathways in MCI conversion involved in inflammation (oleamide, CFP), neuronal regulation (JPH3, SNCA) and protein degradation (PI15).