Team:USTC-Software/Implementation
Implementation
Goals
Implement an integrated and easy-to-use tool platform to help researchers shorten their experimentation and advance their projects.
Design a friendly learning platform for beginners in machine learning to help interested researchers or students learn about basic knowledge about machine learning.
Efforts
A convenient tool
We have conducted extensive paper research to learn about the current development of machine learning in synthetic biology, and have trialed some publicly available machine learning models in person. In addition to this, we also tried to get in touch with the developers of some models and tried our best to investigate the principles of these models in order to expect accurate results to the users give the users accurate results.
We talked with some teachers and students in our school lab to make sure our tool provide the functionality they expect, and also to know how they wanted to interact (with our platform) in order to improve our tool to be more user-friendly.
In the end, we selected machine learning models that provided four properties, subcellular localization, transmembrane topology, secondary structure, and isoelectric sites. We knew that providing just these few models would not satisfy the needs of all researchers, so we designed the hot-plug extensible module to facilitate the addition of more models in the future, which will help more researchers and iGEMers.
A learning platform
We believe that researchers can't solve the problem very well just by relying on other people's tools, and knowing basic knowledge of machine learning will help researchers to choose a model more suitable for their needs.
We went through a lot of introductory materials on machine learning, combined some team members' learning experience about machine learning, and finally we chose some very simple examples to explain the basic principles of machine learning and to avoid confusing beginners. Besides, we also asked teachers for effective ways to deliver knowledge, such as setting up appropriate exercises after learning.
When we used this platform for science activities, we also received feedback from some users. Based on those feedback suggestions. Thus, we modified some layout and improved the interaction. We believe these improvements will make our platform more effective.
Future
In order to bring CAT to more people, we are trying to put CAT on larger related platforms or collaborate with other platforms. We know that CAT has a lot of room for improvement, so we will use our hot-plug extensible module to absorb more models to help more researchers and iGEMers. We will also consider adding scientific and educational knowledge suitable for a wider audience. After all, it’s a saying that wisdom in the mind is better than money in the hand.
Contact Us
Mail:
USTC_Software2021@163.com
University of Science and Technology of China, No.96, JinZhai Road Baohe District, Hefei, Anhui, 230026, P.R.China
