Drug Toxicity

Drug Discovery is the procedure by which new drugs are found. scientists find new medications by using bits of knowledge regarding a disease to design a drug to counter its effects. Drug Toxicity, on the other hand, refers to how harmful a drug/medicine can be. Drug Toxicity occurs when a person intakes too much of a drug which can lead to adverse health effects.

The client faced several issues regarding drug toxicity. The process the client used was time consuming and required clinical human trials to take place before a drug could be approved for production. First, drug discovery, which is a process to discover new candidate drugs, took place. After drug discovery was completed, the search for volunteers for human trials could be conducted, animal trials took place proving to be very crucial. Once a sufficient number of volunteers had enrolled, the clinical trials were initiated. The complete trial was conducted in several phases making the process lengthy and costly. It was only during the clinical trials that the drugs exhibited certain levels of toxicity which could be considered as fatal and have adverse effects on health.

To start off the process, domain knowledge regarding drug discovery and drug toxicity was needed. A team of dedicated individuals was made to take up the task. Once the team had completed their study, research was conducted in order to find out which drugs failed their clinical trials for having high levels of toxicity. A list of all such drugs was compiled and their chemical and target properties were studied and jotted down. The trends of the chemical and target properties of failed drugs were studied and then compared with the chemical and target properties of drugs that had passed the clinical trials. The comparison helped figure out which features contributed to a drug being toxic or not.

A Machine Learning (ML) model was trained on the shortlisted features and the model achieved an accuracy of 88% in the prediction of drugs having high levels of toxicity.

ByteCorp worked closely with the client to use Artificial Intelligence (AI) to predict whether a drug will end up having high levels of toxicity in the future. The solution provides in the future before the clinical trials take place. This allows the complete process to be time and cost efficient.

ByteCorp recommended and implemented an innovative solution based on a Machine Learning (ML) approach, that works towards the prediction of high levels of toxicity in drugs before they can bee sent for clinical trials. The quality control regulation of the drugs makes it more time and cost efficient as drugs containing high levels of toxicity are not cleared for clinical trials. A team consisting of Data Scientists, having experience of around 2.5 years, Doctors and Microbiologist was deployed to take forward the project.

The innovative solution enabled the client to predict whether a drug would show high levels of toxicity in the future before the clinical trials took place. This meant that if a drug was found to be dangerous, it would not be sent for clinical trials and would be discarded. Prediction prior to clinical trial had a direct impact on the overall cost of the process. Furthermore, overall the pace of drug discovery increased hence, making the solution less time consuming.