Melbourne Scientists Use AI to Tame CRISPR

Last year a ten-month old baby in the US was the first person in the world to have their rare genetic disease effectively cured through the use of CRISPR gene editing technology. But the roll out of CRISPR across a wide range of genetic conditions has been hampered by its inconsistency, and its potential to cause harm to healthy genes. Now a team of Melbourne scientists have used AI to develop a fast and accurate way to keep CRISPR in line.

Described as genetic scissors, CRISPR allows scientists to cut, remove and replace faulty genes with remarkable accuracy.

But despite its power, CRISPR technologies have known safety risks: The active enzyme can linger in cells and cause unintended damage to DNA or RNA - so-called off-target effects - which may trigger harmful mutations in healthy genes. Which is where anti-CRISPRs come in. Derived from the viruses that infect bacteria, they can control the CRISPR machinery and keep it in check. The problem is – they are rare and very hard to find.

Now a team led by Associate Professor Gavin Knott at the Monash University Biomedicine Discovery Institute in collaboration with D. Rhys Grinter at the Bio21 Molecular Science and Biotechnology Institute at the University of Melbourne have published a study in the journal, Nature Chemical Biology, that describes the use of AI technology to develop a highly successful and rapid approach to create anti-CRISPR molecules.

Using AI-accelerated protein design, we rapidly produced functional inhibitors of CRISPR that function in bacterial and human cells".

Dr. Cyntia Taveneau, lead author and protein designer, Monash University

According to Associate Professor Knott, the ability to "design bespoke inhibitors that can keep CRISPR "in line" will contribute to the ongoing development of CRISPR tools in diverse applications across research, medicine, agriculture and microbiology,".

In the last decade of CRISPR research only 118 anti-CRISPR molecules have been identified and according to Dr Grinter, "the discovery of natural inhibitors against clinically relevant targets remains challenging and time-consuming."

"In this study we implemented a rapid approach to anti-CRISPR design that uses AI to create highly accurate and specific anti-CRISPRs, in this case to control the activity of an RNA editor."

The new approach is far more rapid than traditional protein discovery processes, taking 8 weeks from target selection through to hit and lead identification, a significant acceleration towards full scale development for use in the clinic.