Implementation of a genetically tuned neural platform in optimizing fluorescence from receptor-ligand binding...

Area of Science:

• Biotechnology
• Analytical Chemistry
• Computational Biology

Background:

• Biosensing relies on sensitive detection of molecular interactions.
• Microfluidic devices offer platforms for high-throughput analysis.
• Optimizing binding conditions is crucial for signal enhancement.

Purpose of the Study:

• To optimize fluorescence detection of teicoplanin antibiotic binding using a neural network.
• To investigate the impact of incubation and buffer flush times on fluorescence signal.
• To validate a genetically tuned neural network approach against traditional methods.

Main Methods:

• Utilized a genetically tuned neural network platform.
• Employed a microfluidic channel with electrostatically attached teicoplanin.
• Optimized parameters including incubation times and buffer flush duration.
• Applied neural network methodology for fluorescence signal optimization.

Main Results:

• The optimal neural network structure provided an excellent fit for both training (r² = 0.985) and testing (r² = 0.967) data.
• Simulated results were experimentally validated, confirming the neural network's efficiency.
• The proposed method demonstrated superiority over multiple linear regression and standard backpropagation neural networks.

Conclusions:

• Genetically tuned neural networks effectively optimize fluorescence detection in microfluidic biosensing.
• This approach offers a robust and efficient alternative for analyzing molecular binding events.
• The validated method shows significant potential for enhancing antibiotic detection sensitivity.