Automated single-nucleotide polymorphism analysis using fluorescence excitation-emission spectroscopy and one-class...

Area of Science:

• Genetics
• Bioinformatics
• Molecular Biology

Background:

• Automated genotype identification is crucial for genetic analysis.
• Single nucleotide polymorphism (SNP) analysis faces challenges like spectral variation and outliers.
• Accurate classification of homozygotic and heterozygotic samples is essential.

Purpose of the Study:

• To develop a highly automated method for SNP analysis and genotype identification.
• To address challenges in spectral variation, heterozygote spectra variability, outliers, and incomplete allele representation.
• To improve the accuracy and reliability of genetic data analysis.

Main Methods:

• Utilized Taqman reaction for data generation and fluorescence spectroscopy for spectral acquisition.
• Implemented a signal-standardisation technique (piecewise direct standardisation, PDS).
• Employed one-class classifiers based on Principal Component Analysis (PCA) models for homozygote identification, followed by linear combination analysis for heterozygote and outlier classification.

Main Results:

• The developed method effectively handles spectral variations and outliers.
• Accurate identification of homozygotic (wild-type and mutant) and heterozygotic genotypes was achieved.
• Achieved very low false-positive errors and 2-6% overall false-negative errors in SNP analysis.

Conclusions:

• The new automated SNP analysis method provides a robust solution for genotype identification.
• The combination of PDS and PCA models significantly improves classification accuracy.
• This approach offers a reliable and efficient tool for genetic research and diagnostics.