AI shown to successfully predict changes in DNA
AI is now at the forefront of DNA structural analysis thanks to new genomics research from a team at Stanford University, California. The paper, published in Nature Genetics reveals that deep learning AI is now capable of understanding and predicting pattern changes in DNA structures.These findings suggest that AI analysis of DNA structures now surpass human capabilities due to its speed and accuracy. This also means that genomics is likely the next field of research to benefit from developments in artificial intelligence.
READ NEXT: The AI shaking up earthquake research
Before AI-accelerated analysis existed, scientists relied on lab research to conduct analysis on changing DNA structures. It’s costly and time-consuming, with DNA sequences often subject to not one, but multiple changes at a time. Using AI, on the other hand, provides the capacity to analyse DNA much faster.
Researchers taught the AI to recognise the outcomes of particular changes in DNA structures, allowing it to develop an understanding. The end result was that the AI could automatically generate new DNA sequences with specific desirable properties. For instance, it could program how a DNA structure might respond to a particular medicine, or how plant crops might respond to alterations in their genome.
The success of the AI’s ability to understand DNA structures permits scientists to run diagnostics, test vaccinations and review crop growth with ease and efficiency. Mikael Huss, a data scientist part of the research team, describes the developments as “game-changing.”
READ NEXT: What is AI?
Building upon this success, the team has launched an online tutorial for AI experts and genomics researchers to utilise. The tutorial offers an “easy-to-use” playbook, explaining how to harness the power of AI specifically for the purpose of genomics research. It’s an intelligent move, as it can help individual researchers with their own studies, while potentially improving their AI, too. That’s a win-win situation, if you ask me.