Discovery AI at Duke

The Duke Discovery AI initiative advances the integration of artificial intelligence with fundamental biological research across Duke University. Our mission extends beyond applying AI to existing biological problems. Our community of scholars and scientists seek to transform how biological systems are interrogated and engineered across all scales, from molecular to cellular to organism. This rapid iteration amongst deep-learning modeling and prediction, hypothesis generation, and experimental validation accelerates discovery.

The initiative unites computational scientists, biologists, and engineers who recognize that the most transformative advances will emerge not from AI serving as an auxiliary tool, but from the deep integration of computational and biological sciences. Through the development of novel machine learning approaches specifically designed for biological complexity, including protein language models that elucidate sequence-function relationships and algorithms that map neural circuits and microbial communities, we are creating methodologies that enable previously infeasible experimental approaches. Our distinguishing characteristic is the seamless integration of computational modeling and experimental validation. Faculty in Discovery AI are developing AI systems that reveal biological patterns that are undetectable through conventional analyses, actively guide experimental design, and compress discovery timelines. Discovery AI researchers are also committed to training the next generation of scientists who possess expertise in both biological systems and computational methods.