Machine learning approach combined with causal relationship inferring unlocks the shared pathomechanism between...

Area of Science:

• Cardiovascular Medicine
• Infectious Diseases
• Bioinformatics

Background:

• Growing evidence indicates a higher prevalence of Acute Myocardial Infarction (AMI) in patients with Coronavirus Disease 2019 (COVID-19).
• The precise biological mechanisms underlying the association between COVID-19 and AMI remain largely unelucidated.

Purpose of the Study:

• To investigate the shared molecular mechanisms and potential causal relationships between COVID-19 and AMI.
• To identify common differentially expressed genes (DEGs) and develop a predictive model for AMI risk in COVID-19 patients.

Main Methods:

• Acquisition of gene expression profiles for COVID-19 and AMI from the Gene Expression Omnibus (GEO) database.
• Identification of common DEGs between the two conditions.
• Application of machine learning algorithms for diagnostic predictor development and Bayesian networks for causal inference.

Main Results:

• Sixty-one common DEGs were identified between COVID-19 and AMI.
• A diagnostic predictor was developed using 20 machine learning algorithms to assess AMI risk in COVID-19 patients.
• Causal relationships and shared immunological implications were explored, revealing key biological processes in co-pathogenesis.

Conclusions:

• This study presents a novel application of causal inference to elucidate the shared pathomechanism between COVID-19 and AMI.
• The findings provide mechanistic insights into the co-occurrence of these diseases, potentially aiding future preventive and precision medicine strategies.