Inspiration
During a conversation with a friend, he shared his experience working on a program that required him to learn and use PyMOL. PyMOL is a biomedical application used to create 3D models of proteins. He explained how difficult and confusing the coding language was. This discussion inspired me to create a language learning model that can generate code automatically, helping users work more easily with complex tools like PyMOL.
What it does
My LLM takes in user input and prints out 3 data cards. Each card gives information on each protein and allows the user to copy the PyMol code for the model given. With the vast database I connected to my project, the RCSB Protein Data Bank, it is possible to create models of virtually any protein.
How I built it
For my project, I built the backend using Windsurf's IDE and Cascade AI, which lets the system understand what the user types and turn it into working code. Then, I used Claude AI to create the website and design a user interface that’s both unique and easy to use. Using these tools together, I was able to make a project that not only works smoothly but also feels polished and user-friendly.
Challenges I ran into
The biggest challenge I ran into while building my project was getting the database to connect. At first, my model just wouldn’t communicate with the website, and I spent hours trying to figure out the API. After a lot of trial and error, I finally got it working, and now everything runs smoothly and quickly.
Accomplishments that I'm proud of
I was most proud of getting the database to connect to my project and of creating a website design that’s both unique and visually appealing.
What I learned
I learned how to utilize multiple resources, such as Claude AI and Windsurf, to not only create the website, but to also enhance its visuals and performance.
What's next for Protogen
I’d love to create a new model on the website that can actually generate 3D protein structures by itself, instead of just giving PyMOL code.
Log in or sign up for Devpost to join the conversation.