Researchers at the University of California, Irvine have developed a new machine learning framework, dubbed SIGNET, to map gene regulatory networks specific to different brain cells affected by Alzheimer’s disease. The work, published in Alzheimer’s & Dementia: The Journal of the Alzheimer’s Association, represents the most detailed analysis to date of how genes interact to drive the progression of the disease.
The team, led by Min Zhang, professor of epidemiology and biostatistics, and Dabao Zhang, also a professor of epidemiology and biostatistics, utilized SIGNET to move beyond identifying genetic correlations and instead pinpoint causal relationships between genes. This approach, they say, is crucial to understanding the underlying mechanisms of Alzheimer’s, which is projected to impact nearly 14 million Americans by 2060.
“Different types of brain cells play distinct roles in Alzheimer’s disease, but how they interact at the molecular level has remained unclear,” explained 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.”
SIGNET was created by analyzing single-cell molecular data from brain samples collected from 272 participants involved in the long-term Religious Orders Study and the Rush Memory and Aging Project. The framework integrates single-cell RNA sequencing and whole-genome sequencing data to reveal cause-and-effect relationships among genes. The researchers identified causal gene regulatory networks for six major types of brain cells, determining which genes are likely controlling others – a capability lacking in traditional correlation-based tools.
“Most gene-mapping tools can reveal which genes move together, but they can’t tell which genes are actually driving the changes,” Dabao Zhang stated. “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 analysis revealed the most significant gene disruptions in Alzheimer’s disease occur within excitatory neurons, the nerve cells responsible for transmitting activating signals. Examination of nearly 6,000 cause-and-effect interactions showed extensive rewiring within these cells as the disease progresses. Researchers also identified hundreds of “hub genes” – major control centers influencing numerous other genes – that could serve as potential targets for early detection and therapeutic intervention.
The study also uncovered new regulatory roles for genes already known to be associated with Alzheimer’s, such as APP. Analysis showed APP strongly controlled other genes within inhibitory neurons. The team validated their findings using an independent set of human brain samples, bolstering confidence in the biological relevance of the identified gene-to-gene relationships.
According to the researchers, SIGNET’s application extends beyond Alzheimer’s disease, offering a powerful tool for studying other complex conditions, including cancer, autoimmune disorders, and mental health conditions. The research was funded in part by the National Institute on Aging and the National Cancer Institute.
