Drug Discovery and Drug Identification using AI

Abstract

The paper deals with identifying and creating new drugs using AI technique. We are implementing a process using Intel Open VINO toolkit for identification of drugs. With this detection technique we can identify the reactants which are added as drugs and automates the entire flow. We are using Intel OpenVINOtoolkit with custom object detection technique to train the model using the faster Region Based Convolutional Neural Network (R-CNN)method with labeled drugs (organic compounds) which act as Reactants. Using this approach, the entire drug discovery process of clinical trial for the process can be reduced to very small time of 3-4 months (which generally takes 10-12 years) and we can generate simulated drugs to see the behavior and implementation becomes faster. We are also creating a customized dataset of drugs or molecules which are used for identifying the drugs. We are using the Canonical SMILES for the molecules so we will map SMILES with organic compounds being detected.