UC San Diego researchers reported Friday that they had decoded the DNA sequence pattern of the initiator, a core segment that helps start gene activation.
A team led by graduate researcher Torrey Rhyne-Carrigg in Professor James T. Kadonaga's laboratory measured gene-expression activity across about 500,000 versions of the initiator using high-throughput DNA sequencing, then trained machine-learning models on the results. The decoded signature indicated that about 60% of human genes contain the initiator.
The announcement is one point in a longer process involving UC San Diego and Artificial Intelligence. Readers should distinguish an early result or institutional summary from replicated findings, peer-reviewed evidence and a tool or treatment that has proved useful outside the original setting.
Researchers said the model could help predict the effects of initiator mutations tied to disease and support the design of synthetic promoters that turn genes on and off. The study used the Expanse CPU at the San Diego Supercomputer Center and received support from the National Science Foundation and National Institutes of Health.
For UC San Diego and Artificial Intelligence, a large number can describe reach without describing effectiveness. The more revealing questions concern who was included, what was measured, how outcomes changed and whether another team could reproduce the finding.
The headline on UC San Diego and Artificial Intelligence compresses a long process into a single result or award. Behind it are methods, funding decisions, institutional choices and the question of whether a finding or program can be repeated, scaled and made useful beyond the team that produced it.
The next questions about UC San Diego and Artificial Intelligence concern methods, limitations, peer review, replication and access to underlying data or code where appropriate. If the work points toward a practical application, later testing will matter more than the promise of eventual use.