Lard traceability in lard-adulterated starch-based food using qPCR and evaluation of chemometrics and random forest on GC-MS chromatogram

The adulteration of lard in food materials can go undetected or encapsulated, increasing the difficulties in identifying food components within. Therefore, a reliable technique for detecting lard adulteration in starch-based foods is required to protect consumers from potential food adulteration. This research aimed to detect lard in lard-adulterated starch-based food samples using real-time PCR (qPCR) and gas chromatography-mass spectrometry (GC-MS) assisted by chemometrics and random forest. For the DNA-based detection method, CTAB and enzyme-CTAB were used to extract DNA from samples of tapioca starch and wheat flour spiked with different percentages of lard; 0%, 3%, 5%, 10% and 50% (v/w). The application of enzymatic treatment using starch-hydrolysing enzymes (α-amylase and amyloglucosidase) on lard-adulterated wheat flour increased the range of extracted DNA concentration from 1.80 ng/μL - 11.23 ng/μL (CTAB) to 3.60 ng/μL - 17.77 ng/μL (enzyme-CTAB) meanwhile, the DNA concentration from lard-adulterated tapioca starch slightly increased from 0.23 ng/μL - 0.33 ng/μL (CTAB) to 0.20 ng/μL to 0.60 ng/μL (enzyme-CTAB). However, the detection of lard in the samples using real-time PCR was unsuccessful. The application of enzymatic treatment in the DNA extraction protocol was ineffective for lard detection. As an alternative, another method was explored by targeting the fatty acid content of wheat biscuits adulterated with 3%, 5%, 10% and 50% lard using GC-MS assisted by chemometrics and random forest. Chemometric analysis of the GC-MS profiles successfully distinguished the unadulterated wheat biscuits from lard and lard-adulterated wheat biscuits. Principal component analysis (PCA) and hierarchical clustering analysis (HCA) analysis clustered all samples into three distinct groups and further confirmed by the partial least squares – discriminant analysis (PLS-DA) and random forest. The random forest model outperformed PLS-DA with a prediction accuracy of 1.0, proposing C18:3n6 as a biomarker for The adulteration of lard in food materials can go undetected or encapsulated, increasing the difficulties in identifying food components within. Therefore, a reliable technique for detecting lard adulteration in starch-based foods is required to protect consumers from potential food adulteration. This research aimed to detect lard in lard-adulterated starch-based food samples using real-time PCR (qPCR) and gas chromatography-mass spectrometry (GC-MS) assisted by chemometrics and random forest. For the DNA-based detection method, CTAB and enzyme-CTAB were used to extract DNA from samples of tapioca starch and wheat flour spiked with different percentages of lard; 0%, 3%, 5%, 10% and 50% (v/w). The application of enzymatic treatment using starch-hydrolysing enzymes (α-amylase and amyloglucosidase) on lard-adulterated wheat flour increased the range of extracted DNA concentration from 1.80 ng/μL - 11.23 ng/μL (CTAB) to 3.60 ng/μL - 17.77 ng/μL (enzyme-CTAB) meanwhile, the DNA concentration from lard-adulterated tapioca starch slightly increased from 0.23 ng/μL - 0.33 ng/μL (CTAB) to 0.20 ng/μL to 0.60 ng/μL (enzyme-CTAB). However, the detection of lard in the samples using real-time PCR was unsuccessful. The application of enzymatic treatment in the DNA extraction protocol was ineffective for lard detection. As an alternative, another method was explored by targeting the fatty acid content of wheat biscuits adulterated with 3%, 5%, 10% and 50% lard using GC-MS assisted by chemometrics and random forest. Chemometric analysis of the GC-MS profiles successfully distinguished the unadulterated wheat biscuits from lard and lard-adulterated wheat biscuits. Principal component analysis (PCA) and hierarchical clustering analysis (HCA) analysis clustered all samples into three distinct groups and further confirmed by the partial least squares – discriminant analysis (PLS-DA) and random forest. The random forest model outperformed PLS-DA with a prediction accuracy of 1.0, proposing C18:3n6 as a biomarker for lard in discriminating unadulterated and lard-adulterated wheat biscuits based on their abundance which was proportionately affected by the increment of lard added. The outcomes of this study can serve as preliminary information in halal authentication to determine lard adulteration in starch-based food products.

Text NUR INANI BINTI AZIZAN - IR.pdf Download (2MB)

Actions (login required)

View Item