81. Prediction of plant based EGFR inhibitors against breast cancer (EGFR) using machine learning model

Aly Abdelkareem

Abraham Peele Karlapudi

I am Dr. Abraham Peele and I have completed my Ph.D. in Biotechnology from Vignan’s Foundation for Science Technology and Research, India. My Ph.D. The thesis Title was Characterization and Emulsifying Activities Of A Quorum Sensing Biosurfactant Produced By A Marine Bacterium. I have completed a Master of Technology in Biotechnology at Vignan’s Foundation for Science Technology and Research, India. I am a highly motivated, self-directed individual, qualified, and have knowledge of concepts related to drug discovery and development. Intermediate Python programming experience with high-performance computing and also a strong knowledge of the latest innovations and technology in computational chemistry.

Abstract

Abraham Peele Karlapudi

Vignan’s Foundation for Science, Technology and Research, Guntur, Andhra Pradesh, India

Breast cancer is a major health concern globally, and identifying effective inhibitors is critical for developing targeted therapies. Epidermal Growth Factor Receptor (EGFR) is a protein that plays a vital role in the development of breast cancer. In this study, we applied machine learning techniques to predict potential inhibitors for EGFR in breast cancer. We utilized a dataset of compounds and their inhibitory activities against EGFR, as well as molecular descriptors, such as structural and physicochemical properties of the compounds. We used various machine learning algorithms, including Random Forest, Support Vector Machines, and Artificial Neural Networks, to build predictive models. We evaluated the performance of the models using various metrics, including accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve (AUC-ROC). Our results demonstrate that the Random Forest model achieved the best performance. We identified several potential inhibitors based on their predicted inhibitory activity, which could be further tested experimentally. Our study highlights the potential of machine learning to aid in the identification of effective inhibitors for breast cancer and provides a useful framework for future drug discovery efforts. The selected plant compounds are screened virtually using the machine learning model, followed by molecular docking and simulation study for the identification of top hit compounds as inhibitors of EGFR. The web-based platform developed using Streamlit for large scale prediction of EGFR inhibitors was deployed using the Heroku cloud application platform.