Combining serum microRNAs and machine learning algorithms for diagnosing infectious fever after HSCT

Area of Science:

• Biomedical Science
• Genomics
• Infectious Disease Diagnostics

Background:

• Infection following hematopoietic stem cell transplantation (HSCT) is a major cause of mortality.
• Fever is a critical indicator of infection, but current diagnostic methods are limited.
• Early diagnosis and treatment of infectious fever post-HSCT are crucial for reducing mortality.

Purpose of the Study:

• To investigate the potential of serum microRNAs as biomarkers for diagnosing infectious fever after HSCT.
• To develop a diagnostic model combining microRNA expression and machine learning algorithms.

Main Methods:

• Serum samples and clinical data were collected from 181 HSCT patients.
• Over 80 infectious-related microRNAs were selected and quantified using quantitative PCR (Q-PCR).
• Random Forest (RF) algorithms were employed to construct a diagnostic formula based on microRNA expression.

Main Results:

• Unsupervised clustering revealed a strong correlation between microRNA expression patterns and infection occurrence.
• A diagnostic formula combining over 10 serum microRNAs achieved a diagnostic accuracy exceeding 0.90.
• Correlations between microRNAs, immune cells, inflammatory factors, pathogens, and prognosis were analyzed.

Conclusions:

• Serum microRNA detection combined with machine learning offers a promising approach for diagnosing infectious fever post-HSCT.
• This method has the potential to significantly improve early detection and management of infections in HSCT patients.
• Further research can validate these findings and translate them into clinical practice.