Contribution
Our contributions to the iGEM community span across wet lab methodologies, computational approaches, and human practices, providing innovative tools and frameworks for synthetic biology research.
Wet Lab
Innovative methodologies and system-level contributions to protein interaction studies and directed evolution
Methodological Contribution
We designed a new method to measure the kinetic parameter kon of binding proteins, combining the split luciferase assay with real-time luminescence detection using a plate reader. This setup enables continuous monitoring of protein–protein interaction rates under experimental conditions.
Compared to traditional techniques such as BLI or SPR, which are expensive and less accessible, our method is simpler, more cost-effective, and suitable for iGEM teams. Unlike conventional ELISA, which can only provide equilibrium constants (KD), our approach directly yields kinetic information such as kon, and can be extended beyond protein–protein interactions or kon alone.
In this way, we provide the iGEM community with a low-cost and versatile platform for studying molecular interaction kinetics, opening opportunities for broader applications in biosensors, protein engineering, and synthetic biology.
| Method | Cost | Accessibility | Kinetic Data | Real-time Monitoring | iGEM Friendly |
|---|---|---|---|---|---|
| Our Method | Low | High | kon directly | Yes | Yes |
| Traditional ELISA | Low | High | KD only | No | Yes |
| BLI/SPR | High | Low | Full kinetics | Yes | No |
System-Level Contribution
We constructed a simplified version of the EcORep system, retaining its core mutagenic function while integrating a controllable switch to regulate activity. This design allows users to turn orthogonal replication on or off according to experimental needs, and significantly reduces the complexity of the original system, making it more accessible for iGEM teams who wish to apply in vivo directed evolution.
To ensure practicality, we also performed experimental validation, measuring both the switching time and the mutation accumulation rate under different conditions. By providing benchmark data and a streamlined system, we aim to offer future iGEM teams a more practical, tunable, and easier-to-adopt EcORep framework for continuous evolution applications.
| System | Complexity | Plasmid Count | Interactions | Controllability | iGEM Adoption |
|---|---|---|---|---|---|
| Our Simplified EcORep | Low | Minimal | Streamlined | High | Easy |
| Traditional EcORep | High | Multiple | Complex | Limited | Difficult |
Deep Learning-Based Iteration Planning
We addressed the challenge of iteration planning in directed evolution by developing a deep learning–based diffusion model that estimates the fraction of random mutants likely to outperform the template. Based on these predictions, we designed a data-driven framework to determine the number of error-prone PCR iterations and sample sizes in each round.
This approach improves efficiency by reducing unnecessary mutations and screening, and enables dynamic planning by providing stage-dependent strategies tailored to different phases of evolution. In doing so, we transform iteration planning from guesswork into a reproducible, model-supported process.
By sharing both the framework and its underlying logic, we provide other iGEM teams with a ready-to-use conceptual tool that can guide the design of more efficient and rational directed evolution experiments. This lowers the entry barrier for teams with limited resources and allows them to spend less effort on trial-and-error, focusing instead on building innovative applications.
Error-Prone PCR Simulator: Predicting Stop Codon Risk
To address the issue of premature stop codon accumulation during iterative rounds of error-prone PCR, we developed a computational simulation based on mutation rates reported in literature.
This model estimates the probability of stop codon emergence per generation, allowing users to visualize how random mutations progressively increase the risk of nonfunctional protein sequences as the number of evolutionary cycles grows.
Our contribution provides a quantitative framework to guide the design of error-prone PCR experiments—helping teams determine the optimal number of mutagenesis rounds before usable gene integrity is compromised.
By integrating real experimental parameters such as base substitution rate and sequence length, this tool helps researchers predict mutation saturation thresholds, avoid nonproductive variants, and improve the efficiency of directed evolution workflows.
Dry Lab
Computational tools and frameworks for protein design and directed evolution
Data-Free In-Silico Directed Evolution
With this, future iGEM teams can
- start from a single structure and generate thousands of plausible variants via RFdiffusion + ProteinMPNN.
- label candidates in silico using fast physics (APBS + Brownian dynamics) to estimate encounter rates and rank designs.
- select the top variants as "next-round samples," enabling iterative, dataset-free directed evolution.
AI Optimizer for Protein Sequences
We have implemented the majority of this workflow during our iGEM season. With this, future iGEM teams can potentially
- train a latent-space autoencoder that fuses ESM sequence embeddings and graph/structure embeddings, then learns a performance head from a labeled library (assays or BD labels).
- optimize in latent space (gradient/BO/search) to propose improved sequences and decode them back to FASTA—ready for synthesis and testing.
- start faster using our pretrained SpyCatcher–SpyTag model as a template (weights + configs), and swap in their own library to fine-tune.
Active-Training Directed Evolution
With this, future iGEM teams can
- close the loop between a pretrained generative model and molecular simulation by actively querying the model for informative sequences, labeling them in silico, and feeding them back for fine-tuning.
- improve model performance with minimal new labels using uncertainty/novelty criteria (e.g., acquisition scores in latent space) to pick the next batch of sequences.
- iterate to better candidates: pretrained → propose → simulate/score → fine-tune autoencoder/decoder ("latent optimizer") → propose again, yielding steadily improved designs.
Seed & Seek: A Completed Protein Optimizer Pipeline
With this brand new pipeline, future teams can be users, generating a user-defined protein!!
- use an end-to-end implementation that covers Seeding (RFdiffusion + ProteinMPNN generation with APBS/Browndye labeling) and Seeking (latent autoencoder optimizer + active training) to propose ranked candidates.
- deploy quickly with our turn-key assets: containers, starter configs, scripts, example notebooks, and an "Implementation" guide on the wiki for reproducible runs on local/HPC/cloud.
- request a full-service run: we adapt objectives/constraints to your target, execute the pipeline, and deliver a reproducible repo plus a shortlist of sequences ready for synthesis.
Human Practice
Fostering global collaboration and innovation through TurBioHacks 2025
TurBioHacks 2025: International Synthetic Biology Hackathon
We have organized TurBioHacks 2025, an international synthetic biology hackathon co-hosted with partners from Stanford, NUS, and IIT Madras, to catalyze global collaboration and innovation. Over a 48-hour sprint, students from high school to university tackled real-world challenges in biotechnology, ranging from oncology drug discovery to protein design, astrobiology, and sustainable biomanufacturing.
Our Key Contributions:
- Multi-track challenge system: Enabling participants to choose focused domains such as drug discovery, cancer diagnostics, and protein optimization, fostering diverse problem-solving approaches.
- Integration of cutting-edge computational tools: Including AlphaFold, RFdiffusion, and generative AI, providing participants with direct access to state-of-the-art protein design pipelines.
- Open collaboration platform: Where over 300 participants worked in cross-disciplinary teams, learning how to turn raw ideas into working prototypes within two days.
- Mentorship and keynote sessions: From academic and industry experts, bridging the gap between student teams and professional researchers.
- Deliverables beyond the hackathon: Such as reproducible GitHub repositories, documentation, and outreach materials, ensuring the projects remain accessible and useful to the broader iGEM community.
- Education and engagement: With resume booklets, certificates, and public showcases that empowered participants to connect with future collaborators, mentors, and sponsors.
This initiative represents our commitment to democratizing access to synthetic biology tools and fostering the next generation of innovators in the field. By bringing together diverse perspectives and providing hands-on experience with cutting-edge technologies, we aim to accelerate the translation of synthetic biology research into real-world applications.
2025 NTSEC Summer Camp × NTHU: Synthetic Biology × AI Cross-disciplinary Innovation Camp
We organized the 2025 NTSEC Summer Camp × NTHU, a two-day program co-hosted with the National Taiwan Science Education Center (NTSEC) to introduce synthetic biology and artificial intelligence to junior and senior high school students through creative, hands-on learning. The camp adopted a "students-teaching-students" model, where our high school interns served as mentors, bridging generational learning and transforming complex science into engaging experiences.
Our Key Contributions:
- Cross-generational mentorship model: High-school interns became mentors for younger participants, transforming their learning from the NTHU Summer Internship into practical teaching experience—building both confidence and communication skills across age groups.
- Interactive learning through creativity: Students explored synthetic biology and AI using storybooks, board games, and experiments, making abstract scientific ideas playful, tangible, and intuitive.
- Integration of AI and experimental science: By combining hands-on wet-lab demonstrations (e.g., bacterial staining and microscopy) with data-driven modeling (MATLAB and machine learning), we showcased how interdisciplinary thinking can bridge biology and computation.
- Mutual learning ecosystem: The camp fostered an environment where students, mentors, and organizers learned from one another, creating a cycle of curiosity and inspiration that extended beyond the event.
- Sustainable education model: This program built a replicable framework linking research, teaching, and outreach, ensuring that future cohorts can continue to expand synthetic biology education across schools and communities.
Original Children's Book Created By Our Team — Panda's Treasure Hunt
We created Panda's Treasure Hunt, an original children's picture book co-designed with our NTHU iGEM Summer Interns to introduce scientific thinking through storytelling.
The book transforms synthetic biology concepts into an imaginative adventure of curiosity and discovery—inviting children to learn science through play, imagination, and participation.
Our Key Contributions:
- Interactive Science Storytelling: The book presents biological ideas such as molecular matching and teamwork through an engaging narrative, allowing young readers to learn by following Panda's adventure of problem-solving and exploration.
- Integration of Design and Education: Developed in parallel with our custom-designed board game, the picture book forms a cohesive learning system where children can read, imagine, and then play, reinforcing learning through creative interaction.
- Participatory Learning Design: Each page includes blank spaces for children to draw or write, turning passive reading into active engagement—encouraging them to think like scientists and co-create the story.
- Empathy-Driven Education: The book was co-designed and tested with children and parents, using bilingual feedback forms that helped refine both content and visual storytelling to better fit learners' perspectives.
- Outreach and Exhibition Impact: Panda's Treasure Hunt was showcased at the 2025 NTSEC Summer Camp × NTHU, NTHU Kindergarten, Zhongxiao Elementary School, and the iGEM 10th Anniversary Exhibition, inspiring children to see science as something they can explore, question, and imagine.
This initiative reflects our belief that science education should begin with imagination. By combining storytelling, art, and scientific thinking, Panda's Treasure Hunt helps children realize that learning biology isn't about memorizing facts—it's about curiosity, creativity, and discovery.
Team Custom-Designed Board Game
We co-designed an original, education-first board game with our NTHU iGEM Summer Interns to translate protein concepts into playful, cooperative problem-solving.
The game extends our team's picture book into hands-on learning—where children read the story, then become the explorers who pair proteins, decode clues, and work as a team.
Our Key Contributions:
- Story-Linked Learning System: Seamless bridge from our children's picture book to gameplay, so learners move from narrative curiosity to interactive practice in one coherent set.
- Concrete Model of Proteins: Core mechanic—pairing complementary markers—represents molecular matching (e.g., SpyTag/SpyCatcher) through simple, tactile actions that young players can intuitively understand.
- Co-Design with Interns + Rapid Iteration: Interns co-created rules, visuals, and playtests. We used bilingual feedback forms to refine designs—such as adding icon-based keyword cards to improve accessibility for early readers.
- Industry-Supported Making: Partnered with Inn 3D Studio to produce durable, high-contrast 3D-printed components, enhancing usability, realism, and classroom readiness.
- Inclusive Teamwork Mechanics: Rules highlight communication, cooperation, and reasoning, fostering soft skills and scientific collaboration alongside conceptual learning.
- Proven Outreach Impact: Showcased and play-tested at the 2025 NTSEC Summer Camp × NTHU, NTHU Kindergarten & Zhongxiao Elementary, and the iGEM 10th Anniversary Exhibition, engaging audiences across age groups.
- Ready to Scale: Planned open-source printables, a digital edition, and expansion packs (e.g., enzyme–substrate sets) so schools and iGEM teams worldwide can adopt and adapt the game.
By turning molecular ideas into a cooperative adventure, our board game makes synthetic biology approachable, memorable, and fun—offering young learners a joyful first step into scientific thinking.