AI meets synthetic biology: New method accelerates DNA design for therapies

In a landmark feat at an intersecting space of artificial intelligence (AI) and human genetics, researchers at Rice University have made a significant advancement in DNA design for medical and biotechnology applications, using AI to tackle one of synthetic biology's greatest challenges: scalability.

For years, programming cells for disease detection or cancer treatment remained hindered by the complexity of figuring out the correct DNA sequences needed for desired cellular behaviours.

Caleb Bashor, a scientist at Rice, noted that finding the right DNA sequence is like searching for a needle in a haystack due to the vast number of possible designs.

Unwavered by the severity of the challenge, the team has introduced a method called CLASSIC, which integrates machine learning with extensive DNA design libraries, enabling rapid predictions of effective genetic circuits before physical testing.

CLASSIC enables the simultaneous creation and analysis of hundreds of thousands to millions of DNA designs, far exceeding previous capabilities.

By mapping DNA sequences to corresponding cellular behaviours, researchers used human embryonic kidney cells engineered to glow when specific genes were activated, indicating stronger genetic activity.

The breakthrough was made possible by combining long-read and short-read sequencing methods, which paved the way for comprehensive data collection.

This innovative approach helped produce massive datasets that trained AI models to understand the relationship between DNA structure and cellular behaviour, which is crucial in achieving perfect predictive accuracy in tests.

The findings show that many genetic designs can effectively achieve similar results, offering flexibility in building resilient biological systems.

This combination of high-throughput DNA design and AI modelling could significantly boost the development of cell-based therapies and other synthetic biology applications, signalling an incredible attainment in DNA design.