Diagnostic host gene signature for distinguishing enteric fever from other febrile diseases

Area of Science:

• Genomics
• Infectious Diseases
• Machine Learning

Background:

• Misdiagnosis of enteric fever leads to severe patient mismanagement and antimicrobial misuse.
• Accurate diagnosis is crucial for effective treatment and understanding global disease burden.

Purpose of the Study:

• To develop a machine learning algorithm for accurate enteric fever diagnosis using host gene expression.
• To validate the diagnostic utility of the identified gene expression signature in a clinical setting.

Main Methods:

• A machine learning algorithm was applied to host gene expression profiles.
• A diagnostic signature was identified to differentiate enteric fever from other febrile illnesses.
• Quantitative PCR (qPCR) was used to validate the expression signatures.

Main Results:

• The diagnostic signature achieved an area under the receiver operating characteristic curve >95%, distinguishing enteric fever with high accuracy.
• The signature identified an additional 12.6% of culture-negative suspected enteric fever cases as likely true infections in Nepal.
• Validated expression signatures show potential as PCR-based diagnostics.

Conclusions:

• Data-driven host gene expression analysis offers a powerful tool for diagnosing febrile illnesses like enteric fever.
• The identified signature can improve diagnostic accuracy, particularly in resource-limited and endemic settings.
• This approach supports better patient management and more accurate disease burden estimations.