This year, we are implementing a cross-disciplinary project, mixing the knowledge from different fields to coordinate them. Since we are constructing a compatible platform this year, we have designed and created modules, from wet-lab to dry-lab, which could be useful for future iGEM teams. In the process, we have met several problems and gained valuable experiences dealing with them. We have built general and delicate models, software, and parts. Here is a short summary of our contribution:
Model: predictions of productivity and efficiency, evidence of inhibition induced by CRISPRi, estimation of time Software: an NLP-based search tool for iGEM standard parts, Ask Lantern. Parts: part collection including a series of serine integrases and their specific recognition sites, constitutive and inducible promoters, and so on.
Our model comprises four main components:
We have developed a powerful search tool, Ask Lantern, specifically designed for iGEM standard parts. Leveraging advanced models like Llama3 and BERT, we trained Ask Lantern using comprehensive data from parts.igem.org. This tool maps part names to their functional meanings, enabling users to search efficiently using natural language descriptions instead of traditional keywords. Ask Lantern enhances search efficiency, helping users quickly find the optimal BioBrick from the extensive database. Our wetlab team has validated its effectiveness, and we believe it will greatly assist other teams in their BioBrick exploration. We also aim to inspire further innovation in applying large language models in software projects. In addition, we have developed a bioinformatics circuit visualization tool to enhance the application of our project.
This year, our project includes a series of serine integrases and their specific recognition sites, expanding the basic parts of iGEM. Our composite parts include some inducible promoters, providing more possibilities for gene pathway design. Our parts collection combines many integrase recognition sites, constitutive and inducible promoters, CRISPRi, and quorum sensing elements. These elements interact to form an exquisite, controllable logic system. We verified the control system's feasibility through experiments and Model. We hope these parts can be further explored and improved, contributing to the synthetic biology community.