Our PROTEUS project has made extensive contributions to the field of synthetic biology, particularly within the iGEM. A detailed overview of these contributions is provided below.
PROTEUS Design Platform
PROTEUS introduced a user-friendly web application and open-source code that fully encapsulates our computational pipeline. This platform enables other iGEM teams and researchers to input target protein sequences and obtain mutation suggestions generated by the PROTEUS software with a single click.
By democratizing access to advanced protein engineering tools, PROTEUS bridges the gap between cutting-edge research and practical application. Its dual-mode interface ("Freshman" one-click mode and "Expert" advanced mode) ensures operability for users with varying levels of expertise, supporting efficient candidate prioritization and experimental tracking.
Methodological Framework
PROTEUS provides a comprehensive and reproducible implementation scheme for integrating protein language models into the DBTL cycle. The project documented the operational workflows for data curation, comparative fine-tuning, and systematic sequence generation, offering a clear technical roadmap for other teams addressing their own protein engineering challenges.
By sharing common issues and corresponding strategies, PROTEUS provides practical insights for researchers in the field to enhance the reproducibility and reliability of their studies.
Open Data and Models
PROTEUS publicly shared a comprehensive dataset derived from ProteinGym, which was functionally categorized to support diverse research needs. Furthermore, the trained PROTEUS models are provided as open resources, establishing a benchmark and starting point for future computational biology projects.
These shared resources accelerate research progress by reducing redundant efforts and fostering collaborative innovation.
Engineered Biological Parts
PROTEUS contributed a newly optimized β-lactamase variant to the synthetic biology community as a standardized biological part. This variant exhibits enhanced ampicillin resistance and can serve as a valuable selection marker or functional component in future synthetic biology projects.
The submission is accompanied by detailed experimental metadata, sequence verification, and usage instructions to assist teams in responsibly reusing and adapting the part. Additionally, the part is annotated with biosafety and ethical guidelines to ensure its application aligns with the principles of responsible innovation.
Community Impact
By executing a complete DBTL cycle, PROTEUS demonstrates the deep integration of AI-driven computational methods with wet-lab validation. The detailed tutorials and demonstration datasets included in the project documentation serve as learning tools, lowering the barrier for iGEMers to apply protein language models in synthetic biology projects.
By illustrating how machine learning and experimental biology can synergistically advance protein engineering, PROTEUS inspires a new generation of scientists to explore interdisciplinary research approaches.
Human Practices
Inspired by the DBTL cycle, BIT-LLM proposed the HP-4R cycle. Through conferences, expert interviews, and science popularization activities, the team actively engaged with diverse stakeholders, helping to shape the project's development direction to align with practical needs while ensuring ethical considerations were fully addressed.
Through interactive sessions and public forums, PROTEUS initiated discussions on the societal implications of AI-driven protein design. These dialogues provided crucial perspectives for understanding public concerns and expectations, guiding the project's focus towards responsible innovation.
Educational Outreach
BIT-LLM organized science popularization activities in communities, schools, and welfare institutions, introducing concepts of synthetic biology and artificial intelligence through vivid analogies and interactive sessions. These activities stimulated scientific interest among youth and enhanced public awareness of biotechnology.
Concurrently, BIT-LLM actively participated in outreach events. These activities provided platforms for knowledge exchange, inspiring students and researchers to explore interdisciplinary applications of AI and biology.
Safety and Security Protocols
Beyond conventional laboratory safety sections, BIT-LLM also re-examined the impact of AI on biosafety. After extensive investigation of materials, the team compiled a manual concerning AI and biosafety.
It was also recognized that safety is further reflected in how potential novel societal and ethical risks are proactively identified, managed, and governed. Consequently, safety principles were deeply integrated into every aspect of our Human Practices.