origen

OriGen is a language model capable of generating host-dependent plasmid replicons—the minimal genetic units required for replication. This repository contains the code implementation described in our preprint "Generating functional plasmid origins with OriGen", which demonstrates the first experimentally validated AI-generated sequences capable of biological replication.

This repository includes:

• Model training and inference scripts
• Evaluation pipelines
• Example notebook demonstrating usage
• Example notebook demonstrating pipeline used to generate sequences at varying similarity levels along with matched random sequences, as experimentally validated in paper