Combination of machine learning and protein‑protein interaction network established one ATM‑DPP4‑TXN ferroptotic diagnostic model with experimental validation
Open Access
- 2 July 2025
- journal article
- Published by Spandidos Publications in Molecular Medicine Reports
- Vol. 32 (3) , 1-20
Abstract
Ferroptosis and lethal sepsis are interlinked, although this association remains largely unknown to clinical panels. Sepsis is characterized by dysfunction of the inflammatory microenvironment. Most septic biomarkers lack independent validation, and a comprehensive diagnosis comprising biomarker assessment combined with clinical evaluation may improve sepsis management. Targeting ferroptosis regulators may offer new hope for uncovering the inflammatory machinery and for developing novel diagnostic methods for sepsis, and bioinformatics analyses are a valuable tool to investigate this further. In the present study, septic datasets were obtained from the Gene Expression Omnibus database. Differentially expressed genes (DEGs) were subsequently introduced in enrichment analyses and intersected with ferroptotic genes for acquiring ferroptosis‑related DEGs (FRDEGs). A protein‑protein interaction network (PPIN) was then constructed to retain hub‑FRDEGs, and this was imported into three machine learning algorithms. A nomogram based on the logistic regression model was subsequently built and validated in silico. CIBERSORT and single‑sample gene set enrichment analysis were used to carry out an analysis of the immune microenvironment, and inflammatory associations with the hub‑FRDEGs were examined. A cellular model was subsequently applied to substantiate the results of the bioinformatic analyses. A total of 94 FRDEGs were obtained from the overlap of 4,410 DEGs and 506 ferroptotic genes. One PPIN of FRDEGs was constructed to identify 38 hub‑FRDEGs, and the three machine learning algorithms were subsequently analyzed, which resulted in the identification of three hub‑FRDEGs, namely ataxia telangiectasia mutated, dipeptidyl peptidase 4 and thioredoxin. One diagnostic nomogram was advanced and scrutinized for its diagnostic accuracy. The functions and pathways of the DEGs were revealed to be mainly concentrated on the immune response and cellular transportation. A notably wide discrepancy was demonstrated to exist between the hub‑FRDEGs and the immunocytes. In conclusion, three potential hub‑FRDEGs connected with sepsis were identified in the present study. Their diagnostic accuracy and immune association demonstrated that ferroptosis is implicated in the inflammatory dysfunction of sepsis, and based on these findings, novel strategies for pharmacological interference and improving diagnostic utility may be developed to facilitate improved management of sepsis.Keywords
This publication has 98 references indexed in Scilit:
- Dipeptidyl peptidase 4: A predictor of ferroptosis in ulcerative colitisThe Journal of Gene Medicine, 2024
- Machine learning reveals ferroptosis features and a novel ferroptosis classifier in patients with sepsisImmunity, Inflammation and Disease, 2024
- Identification of RRM2 as a key ferroptosis-related gene in sepsisInflammation Research, 2024
- How to use biomarkers of infection or sepsis at the bedside: guide to cliniciansIntensive Care Medicine, 2023
- Profiling Tumor Infiltrating Immune Cells with CIBERSORT2018
- The GeneCards Suite: From Gene Data Mining to Disease Genome Sequence AnalysesCurrent Protocols in Bioinformatics, 2016
- The Third International Consensus Definitions for Sepsis and Septic Shock (Sepsis-3)JAMA, 2016
- Gliptin and GLP‐1 analog treatment improves survival and vascular inflammation/dysfunction in animals with lipopolysaccharide‐induced endotoxemiaBasic Research in Cardiology, 2015
- Statistics for High-Dimensional Data2011
- Analysis of Relative Gene Expression Data Using Real-Time Quantitative PCR and the 2−ΔΔCT MethodMethods, 2001