UC San Diego researchers use AI to decode key DNA initiator sequence

Researchers at the University of California San Diego have utilized machine learning to decode the DNA sequence of the ‘initiator,’ a critical segment responsible for starting gene activation. By analyzing approximately 500,000 different versions of the initiator using high-throughput DNA sequencing, the team developed an AI model capable of identifying the signature DNA pattern that dictates whether a gene is expressed.

The study, led by graduate researcher Torrey Rhyne-Carrigg in Professor James T. Kadonaga’s laboratory, revealed that roughly 60% of human genes contain this initiator. This breakthrough allows scientists to predict how specific mutations in these sequences might lead to various disorders, including cancer, by disrupting the precise orchestration of genetic instructions. Additionally, the findings could facilitate the design of synthetic promoters to control gene activity.