Examining the Therapeutic Efficacy of Glucocorticoids using Machine Learning Based Transcriptomic Target Profiling in Autoimmune Diseases

Abstract

Immune system protects human body from infections or foreign entities through a well organised mechanism but sometimes autoimmunity is activated. Autoimmune disorders such as posriasis, multiple sclerosis, systematic lupus erythematosus, and addison’s diseases are majorly caused by attack of autoantigens (T cells and B cells) on healthy cells of body. Autoimmunity can be activated due to various factors including genetic factors (STAT4, PTPN22), environmental factors (pollutants), nutritional factors (vitamin D deficiency), and Smoking. Some of the treatments for autoimmunity include immunsuppresive drugs, orally administered autoantigens, Haematopoietic Stem Cell Treatment(HSCT). The most common treatment include intake of corticosteriods such as dexamethasone, beclomethasone, cortisone, and budesonide. Corticosteroids are found to act an anti-inflammatory agents by suppressing inflammatory factors through various ways. While suppressing inflammatory agents they are also found to be responsible for various side effects including glaucoma, osteoporosis, growth retardation. The purpose of this study is to identify ways to control these side effects by identifying therapeutic targets. For this study differential expression analysis was performed on RNA-seq datasets. Common therapeutic targets were identified by performing comparative analysis on differentially expressed genes of all the datasets used for this study. Pathway analysis was also performed to identify effected pathways. Proteins of common therapeutic targets are selected for further study. These proteins were docked with corticosteroids.Corticosteroids used for the study were those which were commonly used for treatment. Protein-ligand interaction was performed to identify how these steroids interact with proteins. Furthermore, Machine learning pipeline was also generated to predict all the possible interactions.