✔ Competition DeliverablesOur Competition Deliverables consist of the Wiki, Presentation Video and Judging Form. We would like briefly show you how great our project is by all these required deliverables.
✔ AttributionsWe are from different departments, and because of iGEM, we come together and collide ideas. Besides, the success of the project can’t be achieved without the support of some teachers and students. You can read more about those people who have supported us through this project here.
✔ Project DescriptionWe were inspired by a presentation on the emerging intersection of machine learning and synthetic biology, Our biochemistry faculty and classmates solidified our confidence that our project would be successful. See for more details here on how we achieved our project goals, its unique advantages, and more .
✔ ContributionAs a software team, we believe that we have made a substantial contribution to the iGEM community. We provide a tool for researchers and iGEMers to use machine learning to predict some protein properties. Besides we also provide a tool to predict RNA secondary structure, the project developed by SJTU-Software.
Apart from that, we have an educational version, which we’re committed to use to make more students more aware of synthetic biology and machine learning. You can see more detailed information about our contribution for iGEM community and synthetic biology here.
✔ Engineering SuccessOur project went through several iterations, which included the idea of the project, the birth of the education version and the detailed design of the project, etc. With each iteration, we got a little better at our project. Eventually we can provide tools to predict protein properties and more importantly can contribute to education in synthetic biology and machine learning. Here are the details about our engineering progress.
✔ CollaborationWe received a lot of help and support from the USTC team from the very beginning of our design, and with their help we have continued to optimize and improve our project - thanks to the constant online and offline communication and meetings between our two teams. In addition, at the CCiC conference, we met constantly with the SJTU-Software team and exchanged ideas about synthetic biology, machine learning, and other areas. For more details, click here.
✔ Human PracticesOur project can really simplify protein property prediction and help people understand the intersection of machine learning and synthetic biology. We are constantly talking to teachers and students to get their input on our project. We participated in the CCiC 2021 and exchanged with other university teams. We also collaborate with science museums and schools to organize large scale science events. Click Human Practices for more details.
✔ Proposed ImplementationOur project is aimed at synthetic biologists who need to predict protein properties or who are interested in machine learning. With our platform they can retrieve faster or enter the field of machine learning faster. In addition, we have detailed the possible risks and our response in this link.
✔ Integrated Human PracticesWe are constantly reflecting on our projects in human practice, which helps us improve and enhance our platform. Through continuous communication with teachers, our project has evolved from a single machine learning prediction model to a comprehensive platform; through communication with students, we have been able to optimize the presentation of our model to make it more understandable to beginners. At the same time, we collected feedback from users on the beta version and made more complete modifications and adjustments to the platform to get the current version. We have documented our process and click Human Practices for more details.
✔ Project ModellingMany technics are implemented in our web design model, including Nginx for load balance, gunicorn for WSGI middleware, Flask for api design, Redis for result cache. And for biology part, there are also some machine-learning models to be illustrated in detail. You will know more about our biology and software modules here.
✔ Proof of ConceptThe software incorporated in our platform has the support of published papers. The concepts and models used have been tested by various cross-validations and have obtained notable results. We had to ensure that the included software packages have a very high accuracy to facilitate the synthetic biologists’ use. Visit the link for details.
✔ PartnershipIn our ongoing conversations with the USTC and SJTU-Software teams, we realized that our partnership could be deepened. At a later stage of the project, we had joint group meetings with the USTC team to discuss the progress of the project and the optimizations that could be made. In talking with the SJTU-Software team, we realized that we could try to accommodate each other's content. So we kept communicating with them and eventually made their package available to run on our platform to deepen the understanding of the relationship between RNA and proteins. You can know details here.
✔ Education & CommunicationWe have made an educational version to enable more students to understand and learn about synthetic biology and machine learning. We have given a lot of thought to the effectiveness of education, especially in terms of interaction. We use easy-to-understand examples to explain the principles. In addition, we have made a brochure to make the education more effective. We have conducted many educational activities using educational version and brochures, and they all have achieved good results. You can visit this link to know details about our educational work.
We had a lot of communication with the professional teachers at our school to make sure that our program was helpful for synthetic biology. We also talked to the directors of our educational activities to make sure that our educational content and approach was effective for the students. Here are the details about our work on Communication.