Data-driven analysis that integrates bioinformatics and machine learning uncovers PANoptosis-related diagnostic genes in early pediatric septic shock.

Objectives: Sepsis is one of the leading causes of death for children worldwide. Additionally, refractory septic shock is one of the most significant groups that contributes to a high death rate. The interaction of pyroptosis, apoptosis, and necroptosis results in a unique inflammatory cell death mechanism known as PANoptosis. An increasing amount of evidence suggests that PANoptosis can be brought on by several stimuli, including cytokine storms, malignancy, and bacterial or viral infections. The goal of this study is to improve the diagnostic significance of the PANoptosis-related gene signature in early pediatric septic shock. Design and methods: We examined children with septic shock from the GSE66099 discovery cohort and looked at differentially expressed genes (DEGs). To filter the important modules, weighted gene co-expression network analysis (WCGNA) was employed. In the end, random forest analysis and the least absolute shrinkage and selection operator (LASSO) were used to determine the PANoptosis diagnostic signature genes. To determine the PANoptosis signature genes, we also found four validation cohorts: GSE26378, GSE26440, GSE8121, and GSE13904. The area under the curve (AUC) of the receiver operating characteristic curves (ROCs), along with sensitivity, specificity, positive predictive value, and negative predictive value, were used to assess the diagnostic efficacy of these signature genes. Results: From GSE66099, 1142 DEGs in total were tested. Following the WGCNA clustering of the data into 16 modules, the MEgrey module showed a significant correlation with pediatric septic shock (p < 0.0001). Following the use of LASSO and random forest algorithms to identify the PANoptosis-related signature genes, which include ANXA3, S100A9, TXN, CLEC5A, and TMEM263. These signature genes' receiver operating characteristic curves (ROCs) were confirmed in the external dataset from GSE26378, GSE26440, GSE8121, and GSE13904, and were 0.994 (95 % CI 0.987-0.999), 0.987 (95 % CI 0.974-0.997), 0.957 (95 % CI 0.927-0.981), 0.974 (95 % CI 0.954-0.988), 0.897 (95 % CI 0.846-0.941), respectively. Conclusion: In summary, the discovery of PANoptosis genes, ANXA3, S100A9, TXN, CLEC5A, and TMEM263 proved to be quite helpful in the early detection of pediatric septic shock patients. These early results, which need to be further confirmed in basic and clinical research, are extremely important for understanding immune cell infiltration in the pathophysiology of pediatric septic shock.