On this page, we have compiled all medal criteria along with links to the corresponding sections of our Wiki. Our project, Cytopia, centers on the human cathelicidin antimicrobial peptide LL-37. Using synthetic biology, we engineered a Saccharomyces cerevisiae chassis for efficient LL-37 production, and further applied machine learning to predict variants with enhanced antimicrobial activity. We also engaged actively with diverse communities, continuously refining our research plan through mutual exchange, and ultimately achieving engineering success. Throughout our Wiki, we have documented our entire exploration in detail. Our project makes meaningful contributions to the iGEM community, to future researchers working on antimicrobial peptides, and to the public groups we interacted with.
We have completed our Wiki and Judging Form, are about to complete our Presentation Video submission, and we are delighted to have the opportunity to participate in the in-person Judging Session.
On our Attribution page, we used a standardized Attribution Form to fairly and accurately acknowledge each team member's work and contributions to the project. We would also like to express our sincere appreciation to everyone who has supported our project.
Visit our Attributions page for details.
Our Cytopia, developed a Saccharomyces cerevisiae chassis optimized for the biosynthesis of the human antimicrobial peptide LL-37. This chassis also has the potential to express other small-molecule antimicrobial peptides, making it available for use by future iGEM teams.
We further designed the CytoFlow framework, which integrates the CytoEvolve model for generating LL-37 variants with enhanced antimicrobial activity, the CytoGuard model for evaluating antimicrobial peptide efficacy, and the CytoGrow model for analyzing yeast growth and nutrient consumption.
During the project, our Human Practices also contributed to the iGEM community. We compiled and published research on the potential applications of antimicrobial peptides, regulatory compliance, and public perception. These findings not only helped us optimize our project design but also provide valuable guidance for future teams conducting related research and community engagement.
Visit our Contribution page for details.
We successfully completed the DBTL cycles in both wet lab and dry lab. Our engineering consisted of 4 cycles, and through multiple iterations and optimizations, our project was continuously refined, ensuring the rationality of the design and the feasibility of its implementation.
Visit our Engineering page for details.
From the beginning of our project, we deeply integrated human practices into our research process by following an “Explore – Feedback – Update – Validate” cycle. We actively engaged with experts, regulators, industry partners, and the public, gaining critical insights into the cost, regulatory landscape, and public perception of LL-37. These interactions significantly guided our project direction and technical strategies.
Our Integrated Human Practices greatly accelerated project progress—for instance, collaboration with Kangma Biotechnology enabled rapid experimental validation of dry lab-designed LL-37 variants. Moreover, we have contributed to raising broader awareness of synthetic biology and the risks of antibiotic misuse. Looking ahead, we may further explore applications in the pharmaceutical industry.
Visit our Integrated Human Practices page for details.
During the construction of a high-yield LL-37 Saccharomyces cerevisiae chassis, we designed a series of multifunctional composite parts. We built up arm-loxP-URA3-loxP-down arm parts for yeast gene knockout (e.g. BBa_25NP2VZ8), combining universality and stability, making them broadly applicable for future targeted gene deletions. We also assembled parts to assist LL-37 secretion, effectively enhancing its secretion efficiency in yeast (e.g. BBa_25B5P3H6). Additionally, our multi-copy integration parts (BBa_251COFY9) enable stable, multi-copy insertion of target genes into the genome, significantly increasing LL-37 production and providing a versatile tool for efficient expression of other proteins.
All parts used are thoroughly documented in the iGEM Registry. These composite parts not only helped us achieve our goals but also hold significant value for future synthetic biology research and applications due to their versatility and innovation.
Visit our Parts page for details.
We designed the CytoFlow framework, which consists of three models: CytoEvolve, CytoGuard, and CytoGrow, respectively responsible for antimicrobial peptide variant generation, activity prediction with physicochemical property analysis, and host cell growth modeling.
In CytoEvolve, we innovatively combined reinforcement learning with diffusion models as the core of our generation and optimization module. For the specific task of LL-37 variant prediction, the improved variants generated by CytoEvolve demonstrated enhanced antimicrobial activity in experimental validation.
CytoGuard integrates feature extraction modules based on ESM-2, Ankh, and ProtT5, together with a regression model using hypergraph structures and attention-fusion mechanisms, as well as physicochemical analysis. This model enabled us to reduce the number of peptide sequences requiring experimental validation from over 100 to just 4, greatly saving time and cost.
CytoGrow established a system of dynamic equations incorporating substrate consumption, product inhibition, and cell death, successfully fitting yeast growth curves and nutrient consumption profiles. This model allowed us to identify the timing of LL-37 expression in yeast and provided guidance for subsequent experiments and data analysis.
CytoFlow opens a new technical pathway for the rational design and optimization of antimicrobial peptides. This highly integrated approach, combining computation and experimentation, demonstrates strong scalability and also provides inspiration for the development of other bioactive peptides.
Visit our Model page for details.
Guided by the principle that "Knowledge is the outset of action; action is the consummation of knowledge," we implemented a dynamic "Reach – Listen – Adapt – Inspire" educational cycle. Through interactive online tools like the game "Cytopia Defense," collaborative livestreams, hands-on experiments, and public outreach, we engaged over 4,000 participants. We fostered two-way dialogue by actively listening to public feedback on concerns like safety and cost, which directly helped us adapt our communication strategies and even refine our project's direction. All materials are openly shared to inspire a deeper, more responsible public understanding of synthetic biology.
Visit our Education page for details.
