AI Designs Viable 19-Amino-Acid Life

🔑 Enhanced Key Takeaways

• •The research, led by Harris Wang at Columbia University, utilized a 'genome-wide recoding' strategy that required the deletion of the isoleucine tRNA and the corresponding aminoacyl-tRNA synthetase to force the cell to rely exclusively on the AI-designed variants.
• •The study demonstrates that the genetic code is more plastic than previously assumed, as the 19-amino-acid organism maintained growth rates comparable to wild-type E. coli despite the massive structural overhaul of its translational machinery.
• •This achievement serves as a proof-of-concept for 'xenobiology,' providing a platform to incorporate non-canonical amino acids into the proteome by freeing up codons previously occupied by isoleucine.

🛠️ Technical Deep Dive

• •Design Pipeline: Utilized a hierarchical approach where ProteinMPNN was used to design sequences for target structures, followed by AlphaFold2 to predict the folding stability of the redesigned ribosomal subunits.
• •Compensatory Mutations: ESM2 (Evolutionary Scale Modeling) was employed to predict the fitness effects of mutations, ensuring that the redesigned proteins could maintain complex protein-protein interactions within the ribosome complex.
• •Genome Engineering: The team employed MAGE (Multiplex Automated Genome Engineering) to introduce the 382 specific mutations across the 52 ribosomal protein genes simultaneously.
• •Selection Pressure: The strain was evolved in a chemostat for over 450 generations to select for compensatory mutations that stabilized the ribosome assembly process in the absence of isoleucine.

🔮 Future ImplicationsAI analysis grounded in cited sources

⏳ Timeline