Machine learning-assisted Zr-MOF_Fe@SECHA fluorescence and colorimetric dual-mode intelligent platform for sensitive...

Area of Science:

• Biochemistry
• Analytical Chemistry
• Nanotechnology

Background:

• Lipopolysaccharide (LPS) is a critical biomarker for sepsis diagnosis.
• Accurate LPS detection is vital for sepsis prevention and public health.
• Existing detection methods lack sufficient sensitivity and reliability for real-world samples.

Purpose of the Study:

• To develop a sensitive and accurate dual-mode analytical strategy for LPS detection.
• To integrate self-feedback exponential catalytic hairpin assembly (SECHA) with Zr-MOF_Fe for enhanced detection.
• To validate the method's applicability in real samples (human serum and vancomycin) and its potential for machine learning integration.

Main Methods:

• Integration of SECHA with Zr-MOF_Fe for fluorescence and colorimetric dual-mode detection.
• Synthesis of Zr-MOF_Fe with optimized peroxidase activity and fluorescence quenching efficiency.
• Utilizing hairpin and double-stranded DNA probes within the SECHA system for exponential amplification.
• Employing a backpropagation neural network (BPNN) model for intelligent data analysis.

Main Results:

• The Zr-MOF_Fe@SECHA system achieved dual-mode detection with broad linear ranges and low limits of detection for LPS.
• Fluorescence mode: 1 fg mL^-1 to 1 ng mL^-1 (LOD: 0.254 fg mL^-1).
• Colorimetric mode: 5 pg mL^-1 to 5 ng mL^-1 (LOD: 0.143 pg mL^-1).
• The BPNN model enabled accurate LPS concentration determination using fluorescence spectra and colorimetric images.

Conclusions:

• The developed Zr-MOF_Fe@SECHA strategy provides a sensitive, reliable, and selective platform for LPS monitoring.
• The method demonstrates excellent reproducibility and applicability in real biological and pharmaceutical samples.
• This machine learning-assisted approach holds significant potential for early sepsis diagnosis and ensuring drug safety.