Machine Learning Decodes DNA Initiator Pattern Present in 60 Percent of Human Genes
A team at the University of California San Diego has used machine learning to decode the DNA sequence of the initiator, the site where gene expression begins. Graduate student researcher Torrey Rhyne-Carrigg and Professor James T. Kadonaga analyzed gene expression activity across approximately 500,000 different versions of the initiator using high-throughput DNA sequencing technology. They then built an AI model that decoded the initiator's signature DNA pattern. Searching for that pattern, the researchers found about 60 percent of human genes contain the initiator. The AI models now provide strong predictions of the presence or absence of the initiator in human genes. The decoded information gives researchers the ability to predict effects of DNA mutations tied to the initiator that can lead to disorders including cancer. The data and models could also be used to design synthetic promoters with customized functions.
Scientists can now predict DNA mutations in the initiator region that may disrupt gene activation and cause diseases such as cancer. The work also enables design of synthetic promoters that turn genes on and off with tailored properties.
Written by the Genomes desk from the primary source linked above and checked against it. Research use only; not medical advice. Corrections: [email protected].