Machine learning reveals thousands of unknown bacterial immune defense systems

Bacterial immune systems protect against viral invaders called phages by precisely targeting specific phage genetic sequences. This precision has provided inspiration and building blocks for biotechnologies such as CRISPR gene editing systems.

To discover more about these bacterial immune systems, Peter DeWeirdt and colleagues developed a machine learning model called DefensePredictor. This model uses information from a protein’s genetic sequence and the sequences of neighboring proteins in the genome to predict whether a protein is involved in immune defense. When Dewiato Others. We tested DefensePredictor on 69 different strains. Escherichia colithe model predicted hundreds of previously unknown immune systems and validated their protective functions in 42 cases. Analysis of 1,000 bacterial genomes identified approximately 3,000 protein clusters that are distinct from known bacterial immune systems. The researchers made DefensePredictor available to all researchers as an open-source software tool.

Separate research aims to discover additional bacterial immune systems, Ernest Mordrett said. Others. developed a machine learning model to predict antiphage defense systems at scale and applied the model to more than 120 million proteins in bacterial genomes. They identified hundreds of thousands of potential antiphage families. Together, these two studies reveal that bacterial immunity is much more widespread than previously recognized, highlighting how such discoveries can have powerful biotechnological implications.