Ubikwitus

Inspiration

Breast Cancer and Parkinsons disease have been recently linked to a protein called Ubiquitin Hydrolase 1. Ubikwitus is an ai model that aims at driving computational drug discovery for these diseases via targeting of that protein

What it does

It takes in potential drug chemical candidates for the enezyme and, based on an extensive training of the model, makes a prediction of the bioactivity and potency of the compound and helps scientists then decide whether to move on into clinical trials

How we built it

We built it using NVIDIA workbench and google colab while using streamlit for the front end

Challenges we ran into

Training the model took too much RAM and cost us $10 out of pocket

Accomplishments that we're proud of

We have proprietary data that has been properly cleaned and optimized

What we learned

How to multitask jobs and clean data

What's next for Ubikwitus

We will continue gathering more data and improving out model

Built With

• jupyter
• python
• scikit-learn
• seaborn
• streamlit
• workbench