Rhomboid domain containing 2 (RHBDD2): A Potential Biomarker of Sepsis and Septic Shock

Abstract

Sepsis is a serious health situation caused by uncontrolled infection and septic shock is a severe condition of sepsis. Both conditions lead to multi-organ failure and may eventually cause death. Studies have listed it as a major cause of death in the intensive care unit (ICU). This study identified the Rhomboid Domain Containing 2 (RHBDD2) as a potential biomarker of sepsis and septic shock. RHBDD2 is a member of the rhomboid superfamily which are inter membrane serine proteases . RHBDD2 is overexpressed in different types of cancer and associated with estrogen receptor stress and up regulator of estrogen receptor in cancer. It was found out using rats and mice models that constituents of endoplasmic reticulum (ER) stress were upregulated in the hearts of septic animals. Hence, it was considered that due to the involvement of RHBDD2 in ER stress in cancer, it might also play a role in sepsis. By using microarray gene expression data and using different computational techniques this study investigated the role of RHBDD2 in sepsis and septic shock using co-expression analysis and identified the deregulation of RHBDD2 in sepsis using differential expression analysis then verify the role of RHBDD2 with the help of machine learning. Results showed that RHBDD2 is overexpressed in sepsis and septic shock. The gene ontology enrichment analysis using Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and biological functions of RHBDD2 and its co-expressed genes module showed that it is involved in most of the sepsis-related 1 biological functions and involved in many of the infection-related pathways which lead to sepsis and septic shock. It was also found that RHBDD2 has a negatively associated with estrogen receptor in sepsis and septic shock. RHBDD2 regulated by Signal Transducer And Activator Of Transcription 5A (STAT5) and Spi-1 Proto-Oncogene (SPI1) transcriptional factors in sepsis and septic shock. This study also assessed the accuracies of different machine learning algorithms in selecting the best features to identify Sepsis and Septic shock. Top 50 features were selected by features reduction algorithm. The Correlation Based Features Selection (CFS) and Chi-square gave better accuracy than Recursive Features Elimination (RFE) on four classification models Support Vector Machine (SVM), Logistic Regression (LR), Naïve Bayes (NB) and k-Nearest Neighbors(KNN). These features are compared with a module containing RHBDD2 co-expressed genes obtained by co-expression analysis gave more than 50 % genes are the same with CFS and Chi-square algorithms while RFE gave less than 10% similar genes. The identification of the RHBDD2 as a biomarker may facilitate in septic shock diagnosis, treatment, and prognosis.