A study published in Alzheimer's & Dementia: The Journal of the Alzheimer's Association has identified new molecular targets for Alzheimer's disease and other neurodegenerative conditions, using an artificial intelligence platform paired with laboratory testing. The work came from Insilico Medicine, a clinical-stage biotechnology company built around generative AI, in collaboration with researchers at the University of Oslo and Akershus University Hospital led by Dr. Sofie Lautrup and Prof. Evandro Fei Fang.
The central finding concerns what the researchers call the NAD+-mitophagy axis. NAD+ is a molecule cells depend on for energy metabolism, and mitophagy is the housekeeping process by which cells clear out damaged mitochondria. Drawing on large-scale human datasets analyzed with Insilico's PandaOmics platform, the team reports evidence that breakdown in this axis is a central feature of both normal brain aging and neurodegeneration - not a bystander effect.
That distinction matters for where Alzheimer's research goes next. If failing mitochondrial quality control sits upstream in the disease process, then measures of that failure could serve as biomarkers, flagging trouble before cognitive symptoms consolidate, and the machinery itself becomes a candidate for intervention. The authors argue that modulating mitochondrial metabolism has real potential for both diagnosing and treating age-related cognitive decline.
The researchers frame their approach as the point of the exercise as much as the result: human data and AI-driven target discovery, followed by rigorous experimental validation rather than computational prediction alone. What they describe are new directions for biomarker development and future interventions, which is a step well short of a tested therapy.