CELLM: Bridging Natural Language Processing and Synthetic Genetic Circuit Design with AI

Abstract

The complexity of the genetic circuit design limits accessibility and efficiency in synthetic biology. This study presents an integrated system that combines Cello software with large language models (DeepSeek-R1, Phi-4) and the LangChain framework in Python, which allows the creation, analysis, and optimization of genetic circuits using natural language instructions. CELLM automates the translation of textual descriptions into functional designs using Cello v2.1 as the basis for circuit synthesis and LLM for the interpretation of biological requirements and logical optimization. To the best of our knowledge, this work sets a precedent as the first system that integrates language models with synthetic biology design tools such as Cello, demonstrating that natural language processing can be translated into functional biological designs. This approach removes barriers by allowing researchers without bioengineering expertise to prototype genetic circuits using simple instructions.