Combined machine learning identification and experimental validation of NDUFS8 and SUCLG1 as potential biomarkers for mechanical asphyxia

Vollständiger Abstract

Worum geht es in dieser Arbeit?

Introduction Mechanical asphyxia (MA) is a major form of violent death caused by external forces that impair respiration and gas exchange. Despite its frequency in forensic casework, its diagnosis remains challenging because autopsy findings are often non-specific. Methods In this exploratory study, we combined data-independent acquisition (DIA) proteomics with machine-learning methods to identify candidate myocardial markers of MA. A discovery cohort of human left ventricular myocardium was profiled to identify nominally differentially abundant proteins. The leading candidates were confirmed by Western blot and immunohistochemistry and further supported by animal models. Results Two proteins central to mitochondrial energy metabolism, NDUFS8 and SUCLG1, emerged as candidate markers, and both were consistently downregulated in the MA group relative to controls. Discussion These changes point to perturbed mitochondrial bioenergetics and oxidative-stress regulation as features of MA-associated myocardial injury. NDUFS8 and SUCLG1 may thus serve as candidate ancillary markers for mechanical asphyxia.