The AI models were trained on vast amounts of biological data in the form of protein or genome sequences, providing insight into naturally occurring genetic sequences. This information was then applied to tasks such as creating new sequences. The hope is that AI-designed gene-editing tools could be better suited to medical applications than existing CRISPRs. Profluent is looking to partner with companies developing gene-editing therapies to test AI-generated CRISPRs.
Key takeaways:
- Researchers have used a generative AI tool, a protein language model, to design CRISPR gene-editing proteins, some of which have been shown to work as expected in laboratory tests.
- The AI model was trained on millions of protein sequences and microbial genomes to design new CRISPR systems, which include a DNA or RNA-cutting enzyme and RNA molecules that guide where to cut.
- One of the AI-designed CRISPRs, named OpenCRISPR-1, was found to be as efficient at cutting targeted DNA sequences as a widely used bacterial CRISPR–Cas9 enzyme, and it made fewer errors.
- The hope is that AI-designed gene-editing tools could be better suited to medical applications than existing CRISPRs, with potential for partnerships with companies developing gene-editing therapies to test AI-generated CRISPRs.