February 18, 2026

AI platform tracks cause-and-effect gene activity in alzheimer’s patients

alzheimertech
Photo Source: Unsplash.com

Scientists at the University of California, Irvine have charted how genes direct one another inside Alzheimer’s-affected brain cells, shedding new light on how the disease advances.

Min Zhang and Dabao Zhang of UCI’s Joe C. Wen School of Population & Public Health led the research, published in Alzheimer’s & Dementia. Their team developed SIGNET, a machine-learning tool that isolates gene-to-gene control instead of simple correlation.

Alzheimer’s is the leading cause of dementia and is projected to affect nearly 14 million Americans by 2060, placing mounting pressure on families, Medicare and long-term care systems. While genes such as APOE and APP have long been linked to the disease, scientists have struggled to explain how they disrupt normal brain function.

“Different types of brain cells play distinct roles in Alzheimer’s disease, but how they interact at the molecular level has remained unclear,” said Min Zhang. “Our work provides cell type-specific maps of gene regulation in the Alzheimer’s brain, shifting the field from observing correlations to uncovering the causal mechanisms that actively drive disease progression.”

The team analysed single-cell and whole-genome sequencing data from 272 participants in two long-running ageing studies. SIGNET integrates both data streams to construct gene regulatory networks across six major brain cell types.

“Most gene-mapping tools can show which genes move together, but they can’t tell which genes are actually driving the changes,” said Dabao Zhang.

“Some methods also make unrealistic assumptions, such as ignoring feedback loops between genes. Our approach takes advantage of information encoded in DNA to enable the identification of true cause-and-effect relationships between genes in the brain.”

The researchers identified nearly 6,000 cause-and-effect interactions in excitatory neurons and hundreds of hub genes that may serve as future drug targets.

The findings were validated in an independent set of brain samples. Funding support came in part from the National Institute on Aging and the National Cancer Institute.

Subscribe for weekly news

Subscribe For Weekly News

* indicates required