Researchers at the University of California San Diego have used machine learning to precisely decode the DNA fingerprint of the initiator — a molecular switch that sits at the exact point where gene reading begins in roughly 60% of focused human gene promoters — giving scientists a quantitative tool, for the first time, to predict whether specific DNA mutations at that location will disrupt gene activation and potentially cause disease, including cancer, according to the UCSD August 2026 announcement.
The findings, published July 31 in Genes & Development, were led by graduate researcher Torrey Rhyne-Carrigg in the laboratory of Professor James T. Kadonaga in UCSD's Department of Molecular Biology, and represent the latest step in a multi-year effort to build an AI model of the complete human gene expression code — the full set of DNA instructions governing which genes turn on, when, and in which cells. The full study by Rhyne-Carrigg and colleagues is available open access through Genes & Development.
To read more, click here.