The number of new pharmaceutical drugs approved by the FDA has been declining steadily whilst the cost and timeframe required to deliver new drugs to market have exponentially increased. In response to the increasingly difficult task of discovering new drugs for life-threatening diseases, we propose an online platform——that allows individuals to query and devise combinations of unique molecules to serve as a basis for generating novel molecules with desired chemical descriptors.

What it does is a web platform that empowers scientists with tools to generate novel compounds with desired chemical descriptors.

How we built it

With great difficulty. Our team split into our divisions, frontend, backend and A.I and individually built the components, then later worked closely together to join all our relevant components.

Challenges we ran into

Chronic hallucinations induced by the absence of sleep. Installing our machine learning dependencies on a Google cloud VM (literally spent four hours typing pip install over and over again hoping for different results)! and lastly the challenging task of bringing together our individual components and making them work together.

Accomplishments that we're proud of

One of our tasks involved developing our own search engine for our platform. We had to come up with creative ways to tackle this problem, and we're proud of the outcome.

What we learned

We learned a lot working together as a team, much more about installing dependencies on a Google cloud VM and just how tricky it can be to tie an ML algorithm to a front and back end and host it online within 36 hours.

What's next for will focus on refining it's ML and building out the tools of it's platform.

Built With

Share this project: