Thank you for visiting nature.com. You are using a browser version with limited support for CSS. To obtain
the best experience, we recommend you use a more up to date browser (or turn off compatibility mode in
Internet Explorer). In the meantime, to ensure continued support, we are displaying the site without styles
AI predicts the effectiveness and evolution of gene promoter sequences
A long-standing goal of biology is the ability to predict gene expression from DNA sequence. A type of artificial intelligence known as a neural network, combined with high-throughput experiments, now brings this goal a step closer.
Andreas Wagner is in the Department of Evolutionary Biology and Environmental Studies, University of Zurich, Zurich CH-8057, Switzerland, and at the Stellenbosch Institute for Advanced Study, Stellenbosch University, Stellenbosch, South Africa.
Gene expression affects every aspect of life, from the survival of bacteria in specific environments to the anatomy and physiology of the human body. The ability to accurately predict how strongly a gene is expressed on the basis of the DNA sequences that regulate such expression would transform how researchers study biology. But the biochemical machinery that regulates gene expression is tremendously complex, and this goal has eluded biologists’ best efforts for more than 50 years. Writing in Nature, Vaishnav et al.1 take advantage of two key technologies to produce a successful ‘oracle’ for gene expression in the yeast Saccharomyces cerevisiae.