© 2022 SMART. All rights reserved
Here at S.M.A.R.T, our mission is to play a role in creating a knowledge empowering environment to accelerate the discoveries of biological and drug inventions. We hope to continue to expand our database and platform, fully implement S.M.A.R.T, and ensure a widespread and impactful use for all scientists, accelerating drug discovery one at a time.
A neural network developed in PyTorch on the Tox21 dataset. Input a SMILE and our model will predict it's toxicity based on 12 chemical tests.
GET STARTEDThere is a general failure rate of 90% for clinical trials due to undesirable characteristics of the drug, most commonly being toxicity
- Accelerating clinical trial and drug discovery, eliminating toxicity within the byproduct.
- Empowering the scientific community around the world with a machine learning model to understand drug mechanisms to optimize research.
SMART uses machine learning to extract meaningful biological information for any compound, using only the chemical structure. This technology has the potential to expedite early screening in any drug discovery research.
The embedding layer for this project could possibly be used for more than just gauging toxicity effectively. This model could be capable of retrieving information from the SMILES alone.