Compound Screening and Prioritization in Drug Discovery Using Machine Learning

Drug discovery is a technique to uncover and nurture a new medication for several medical conditions and diseases. It entails a range of scientific and research endeavors designed to identify substances that effectively manipulate biological processes to lessen, control, or cure illnesses. Delivering safe and efficient medications that can enhance patients' quality of lives is the ultimate goal of drug discovery. The first and foremost step in the development of drug is compound inspection. Compound screening which is also referred to as High Throughput Screening (HTS) is a component in the drug discovery which identifies and assesses the potential compounds that can be incorporated to contribute to the target compound. We propose a machine learning model named Quantitative Structure-Property Relationship (QSPR) model that can predict the amount of toxicity, solubility, metabolic stability, bioavailability, and the chemical reactivity level when a variety of compounds are combined with each other. This model can determine data from various sources, like chemical databases and already formed compounds, to predict which compounds are most likely to exhibit the desired biological activity and chemical reactiveness. They can also learn from existent data, the active and inactive compounds and also recognize their patterns and characteristics. They can also predict the effect of one compound binding with a specific target. This machine learning model simplifies the drug discovery process by incorporating advanced technological techniques that predicts the potential compound binding, toxicity and reactive level and also the hit identification and approval.