Prediction of feature genes in trauma patients with the TNF rs1800629 A allele using support vector machine

• A total of 133 feature genes were screened out in training set.
• SVM classifier peaked at 100% and 86.2% correct rate in training and validation sets.
• Feature genes were mainly enriched in cell proliferation and ribosome pathway.

BackgroundTumor necrosis factor (TNF)-α variant is closely linked to sepsis syndrome and mortality after severe trauma. We aimed to identify feature genes associated with the TNF rs1800629 A allele in trauma patients and help to direct them toward alternative successful treatment.MethodsIn this study, we used 58 sets of gene expression data from Gene Expression Omnibus to predict the feature genes associated with the TNF rs1800629 A allele in trauma patients. We applied support vector machine (SVM) classifier model for classification prediction combining with leave-one-out cross validation method. Functional annotation of feature genes was carried out to study the biological function using database for annotation, visualization, and integrated discovery (DAVID).ResultsA total of 133 feature genes were screened out and was well differentiated in the training set (14 patients with variant, 15 with wild type). Moreover, SVM classifier peaked in predictive accuracy with 100% correct rate in training set and 86.2% in testing set. Interestingly, functional annotation showed that feature genes, such as HMOX1 (heme oxygenase (decycling) 1) and RPS7 (ribosomal protein S7) were mainly enriched in terms of cell proliferation and ribosome.ConclusionHMOX1 and RPS7 may be key feature genes associated with the TNF rs1800629 A allele and may play a crucial role in the inflammatory response in trauma patients. Moreover, the cell proliferation and ribosome pathway may contribute to the progression of severe trauma.