Discrimination of GMOs Using Terahertz Spectroscopy and CS-SVM

Area of Science:

• Agricultural Science
• Spectroscopy
• Biotechnology

Background:

• Distinguishing genetically modified (GM) and non-GM organisms is crucial for regulatory and consumer purposes.
• Existing methods for GMO detection can be time-consuming, destructive, or require extensive sample preparation.

Purpose of the Study:

• To develop and validate an effective, rapid, and non-invasive identification method for discriminating between GM and non-GM soya seeds.
• To leverage the capabilities of terahertz (THz) spectroscopy combined with advanced machine learning for GMO analysis.

Main Methods:

• Terahertz time-domain spectroscopy (THz-TDS) was employed to acquire spectral data from GM and non-GM soya seed samples across the 0.2–1.2 THz range.
• A Support Vector Machine (SVM) model was utilized for classification, with its key parameters (penalty factor and kernel parameter) optimized using the Cuckoo Search (CS) algorithm, creating a CS-SVM model.

Main Results:

• The CS-SVM model demonstrated high accuracy in distinguishing between GM and non-GM soya seeds based on their THz spectra.
• THz spectroscopy proved effective in capturing unique spectral fingerprints of soya seeds, enabling differentiation.

Conclusions:

• The combination of THz spectroscopy and CS-SVM offers a promising, rapid, reliable, and non-invasive approach for the identification of GM and non-GM organisms, specifically soya seeds.
• This method has the potential to streamline GMO detection processes in agriculture and food safety.