Researchers are leveraging generative artificial intelligence (AI) to design novel CRISPR gene-editing systems, marking a significant advancement in the field. By utilizing a protein language model, a neural network trained on vast protein sequences, scientists have successfully designed CRISPR proteins that demonstrate functionality in laboratory settings. Another team has developed a model trained on microbial genomes to design fresh CRISPR systems, comprising DNA or RNA-cutting enzymes and guiding RNA molecules. These breakthroughs showcase the potential of machine-learning models to create complex gene-editing systems. Ali Madani, a machine-learning scientist, underscores the novelty of designing CRISPR proteins entirely through machine learning, with reported success in editing the human genome. Alan Wong, a synthetic biologist, highlights the limitations of naturally occurring gene-editing systems and the potential of AI-driven approaches to expand the repertoire of CRISPR editors for various applications.

Keywords: AI, CRISPR, gene-editing