Machine learning-assisted ratiometric fluorescence sensor array for recognition of multiple quinolones antibiotics

Area of Science:

• Analytical Chemistry
• Environmental Science
• Biotechnology

Background:

• Simultaneous detection of multiple antibiotic residues is critical for environmental monitoring and public health.
• Existing methods often face challenges in sensitivity, selectivity, and simultaneous analysis.

Purpose of the Study:

• To design and develop a novel dual lanthanide fluorescence probe (GDP-Eu-Tb) for simultaneous detection of quinolone antibiotics (QNs).
• To establish a ratiometric fluorescence sensor array combined with chemometric algorithms for QNs detection.
• To evaluate the performance and applicability of the developed method for antibiotic residue analysis.

Main Methods:

• Design of a dual lanthanide fluorescence probe (GDP-Eu-Tb) based on nucleotides.
• Utilizing the inner filter effect (IFE) for fluorescence quenching by QNs.
• Construction of a ratiometric fluorescence sensor array with principal component analysis (PCA) and decision tree (DT) algorithms.
• Performance evaluation using accuracy, precision, recall, and F1 score with ten-fold cross-validation.

Main Results:

• The GDP-Eu-Tb probe enabled ratio fluorescence detection of levofloxacin (LVLX), gatifloxacin (GTLX), and moxifloxacin (MXLX).
• The combined PCA and DT models accurately correlated fluorescence intensity ratios with QNs concentrations.
• The method demonstrated high sensitivity, selectivity, stability, and generalizability.

Conclusions:

• The developed dual lanthanide fluorescence probe and chemometric approach provide an effective solution for simultaneous antibiotic residue detection.
• This method holds significant potential for environmental monitoring and safeguarding human health.
• The study highlights the utility of lanthanide-based probes and advanced algorithms in analytical chemistry.