New AI Model Examines DNA Methylation to Better Understand Mental Illness
New research from the Lieber Institute for Brain Development uses artificial intelligence (AI) to uncover how small changes in our DNA sequence influence a process called DNA methylation. Methylation is a crucial process that regulates gene expression in different types of brain cells.
In a paper published this week in the journal Science Advances, Lieber Institute researchers describe a powerful deep-learning model that leverages existing DNA methylation data to reveal how certain genetic variants impact DNA methylation. The AI tool also connects these changes to brain disorders such as Alzheimer’s disease and schizophrenia.
“What’s new here is our ability to use artificial intelligence to peel back the mystery of gene regulation at the level of individual brain cell types, rather than just bulk brain tissue, and accurately predict which genetic changes matter,” says Shizhong Han, Ph.D., senior author on the paper and a Lead Investigator at the Lieber Institute for Brain Development.
“This level of precision opens the door to identifying the specific genes and cell types driving complex diseases,” Dr. Han adds.
Dr. Han collaborated on the research with first author Jiyun Zhou, Ph.D., a Staff Scientist II at the Lieber Institute, as well as Institute Director & CEO Daniel Weinberger, M.D., also a senior author on the paper.
“For people living with Alzheimer’s, schizophrenia, and other brain disorders, these findings could pave the way for pinpointing entirely new drug targets,” says Dr. Weinberger.
By identifying the specific genetic and molecular drivers behind these conditions, this approach could enable researchers to develop therapies that directly address the root causes of disease. Ultimately, it could lead to treatments that are more effective, with fewer side effects, and better outcomes for patients.
“This work underscores the potential of artificial intelligence to uncover previously hidden secrets of how cells work in the brain,” Dr. Weinberger says.