What question did this study set out to answer?

This study aims to identify DDR genes linked to acute myocardial infarction and develop a machine learning risk prediction model for the Bangladeshi population.

May 9, 2026Open Access

Integrative transcriptomic and machine learning approach reveals key DDR genes and predictors of ami risk in the Bangladeshi population

Key Points

This study aims to identify DDR genes linked to acute myocardial infarction and develop a machine learning risk prediction model for the Bangladeshi population.
Differential expression analysis and WGCNA were conducted to identify DDR genes associated with AMI.
Key candidates were validated using RT-qPCR in Bangladeshi patients.
A machine learning model was trained using demographic and gene expression data with five classifiers.
OGG1 was confirmed as significantly upregulated by ~2.78-fold in AMI patients (p < 0.001).
NBN expression was significantly higher in smokers (p < 0.05).
The logistic regression model achieved an accuracy of 87.3% and an AUC of 0.903.

Abstract

Acute Myocardial Infarction (AMI) is a significant contributor to cardiovascular death, with rising prevalence in Bangladesh. Although DNA damage contributes to AMI pathogenesis, the role of DNA damage repair (DDR) genes remains poorly characterized. This study aims to identify key DDR genes associated with AMI, and develop a machine learning-based risk prediction model for the Bangladeshi population. Differentially expressed DDR genes in AMI were identified via bioinformatics analysis, with WGCNA revealing AMI-associated hub genes. These candidates were validated by qRT-PCR in Bangladeshi patients. A machine learning model using demographic and gene expression data was then trained using five classifiers. Differential expression and WGCNA identified Nibrin (NBN) and 8-Oxoguanine glycosylase (OGG1) as key dysregulated DDR genes. RT-qPCR confirmed a ~ 2.78-fold upregulation of OGG1 (p < 0.001) in AMI patients, while NBN expression was significantly higher in smokers (p < 0.05). Feature selection identified age, smoking, hypertension, diabetes, BMI, and OGG1 as critical AMI risk predictors. Logistic regression showed the best performance (accuracy 87.3%, AUC = 0.903). This study provides novel evidence linking DDR genes to AMI and presents a strong ML-based risk stratification model for the first time, integrating transcriptomic and clinical data from Bangladeshi patients to advance personalized AMI management.

اسأل الذكاء الاصطناعي

Bookmark

View Full Paper