Researchers at the University of California San Diego used laboratory tests and artificial intelligence to study a small but important part of human DNA called the initiator.The initiator helps mark where the instructions in a gene begin to be read. The team measured gene activity in about 500,000 versions of the sequence and used those results to train a machine-learning model.The model identified a pattern linked to the initiator and found that roughly 60 percent of human genes contain it. Knowing that pattern may help researchers understand how some DNA changes alter gene activity.The work could also help scientists design carefully controlled genetic switches for research. It is one piece of a much larger effort to understand how the six billion DNA bases in a human cell work together.