Histones Classification Based on EGFET Signals.

Area of Science:

• Biomedical Engineering
• Molecular Biology
• Analytical Chemistry

Background:

• Histone dysregulation is linked to cancers and neurodegenerative diseases.
• Histone-specific biosensors are vital for chromatin dynamics and epigenetic regulation research.
• Advancements in biosensor technology can drive breakthroughs in cancer research and personalized medicine.

Purpose of the Study:

• To quantify the biosensor's ability to differentiate Human Histones (H4) from other molecules.
• To analyze biosensor performance using classification methods and machine learning.
• To explore the impact of manufacturing defects and environmental factors on histone detection.

Main Methods:

• Development of a biosensor using a KU7 RNA aptamer on a gold electrode.
• Application of classification methods to analyze biosensor data.
• Utilizing machine learning for exploratory analysis of sensor performance under varying conditions.

Main Results:

• Achieved high classification performance with an F1 score exceeding 0.99 for Human Histones (H4) detection.
• Identified key features providing physical insights into sensor operation beyond typical analysis.
• Machine learning analysis demonstrated robustness against manufacturing variations and environmental changes.

Conclusions:

• The developed biosensor demonstrates high specificity and accuracy for Human Histones (H4) detection.
• The sensor's performance insights can enhance future biosensor design and application.
• Machine learning integration offers a pathway for reliable histone detection in diverse, real-world scenarios.