CHEMOMETRIC EVALUATION ON PROFILES OF LARD AND SELECTED EDIBLE FATS AFTER HEATING-PROCESS USING SPECTROSCOPY AND CHROMATOGRAPHY TECHNIQUES Nor Aishah Binti Mohd. Salleh Thesis submitted in partial fulfilment for the degree of DOCTOR OF PHILOSOPHY IN SCIENCE AND TECHNOLOGY UNIVERSITI SAINS ISLAM MALAYSIA February 2023 ii AUTHOR DECLARATION I hereby declare that the work in this thesis is my own except for quotations and summaries, which have been duly acknowledged. Signature: Date: 24th February 2023 Name: Nor Aishah Binti Mohd Salleh Matric No: 4160238 Address: Lot 2676 Tmn Muji, 22200 Besut Terengganu iii ACKNOWLEDGEMENTS I am grateful to Allah SWT, and His Messenger Muhammad SAW, the only inner strength to finish this research, especially during the Covid-19 pandemic constraints. The most appreciation is to Associate Prof. ChM. Dr. Mohd Sukri Hassan, the main supervisor of this research project, for his continued support and guidance. The second is thanks to co-supervisor, Associate Prof. ChM. Dr. Juliana Jumal, for advice on research and thesis writing. The following thanks to the Deans of the Faculty of Science and Technology (FST), Heads of the Centre Graduates Studies (CGS) Department, International Fatwa and Centre (iFFAH), and FST staff for cooperation in the lab work. I would like to thank the postgraduate students (especially Nurul Elma and Khalilah) for their bitter and sweet memories. Special thanks to my beloved husband (Sharil Nizam), children (Aryana & Adam), and sister (Nurulhuda) for their endless support. Finally, highly appreciative is to the Ministry of Higher Education Malaysia (MOHE) for financial research support of the FRGS-50715 and USIM grant under Centre of Excellence (COE), iFFAH (PPPI /KHAS_IHRAM /04/061007/13818). iv ABSTRAK Tujuan penyelidikan ini adalah menilai profil lemak babi menggunakan kaedah kemometriks untuk mendiskriminasikan lemak babi daripada jenis lemak lain setelah proses pemanasan dan penentuan biomarker dalam lemak babi. Dalam penyelidikan ini, dua jenis persampelan dilakukan. Pertama, sejumlah 270 sampel lemak babi daripada lemak perut (BL), lemak belakang (BK) dan lemak bahu (SF) dikumpulkan dari wilayah utara, selatan dan tengah, Semenanjung Malaysia dan diuji menggunakan teknik FTIR. Penilaian secara kemometriks ke atas profil lemak babi dilakukan menggunakan PCA dan dilanjutkan dengan menggunakan teknik Hotelling T2. Persamaan vektor skor ditemui pada 3PC pertama (99.5% selang keyakinan) dan berjaya disahkan oleh projeksi PCA. Ini menunjukkan profil FTIR bagi lemak babi adalah serupa bagi semua wilayah dan bahagian badan. Kedua, untuk tujuan diskriminasi, sebanyak 60 sampel- sampel lemak dari babi, ayam, lembu, kambing dan tumbuhan telah disediakan menggunakan dua belas protokol proses pemanasan (pada suhu 120 ° C, 180 ° C, 240 ° C selama 0.5, 1, 2 & 3 jam). Lemak-lemak haiwan dan tumbuhan tanpa pemanasan juga termasuk dalam tambahan kajian ini. Selepas proses pemanasan, sampel-sampel lemak dianalisis menggunakan FTIR, 1H-NMR, 13C-NMR, GC-FID dan LC-MS/MS. Diskriminasi lemak-lemak dilakukan menggunakan teknik-teknik pengkelasan multivariasi (LDA, MDA, QDA & SVMDA) dan regresi multivariasi (OSC-PLSR, PCR & PLSR) pada FTIR dan 1H-NMR data. Pengkelasan multivariasi mendapati MDA, QDA, dan SVMDA pada 1H-NMR memberikan keputusan yang terbaik untuk pengelasan lemak babi daripada lemak-lemak lain. Bagi regresi multivariasi, keputusan OSC-PLSR pada FTIR (R2, adj. R2, RMSEC, RMSEV & MSEP; 0.985, 0.984 , 0.049, 0.051 & 0.056) mengatasi prestasi keputusan OSC-PLSR pada 1H-NMR. Akhir sekali, pemprofilan lemak babi oleh 13C-NMR, GC-FID dan LC-MS/MS untuk menentukan biomarker lemak babi telah dilakukan menggunakan teknik PCA. Data 13C-NMR-PCA menunjukkan lemak babi dapat dibezakan dengan lemak ayam pada C-2 oleh isomer TAG yang dikaitkan dengan resonan δ 34.21 dan δ 62.10. Manakala, data GC-FID-PCA mendapati bahawa lemak haiwan dan lemak tumbuhan boleh dibezakan selepas degradasi isomer asid lemak (FAs) cis kepada trans. Lemak babi yang dipanaskan pada 180 °C (pada 0.5, 1 & 2 jam) didapati berbeza lebih ketara oleh LC-MS/MS-PCA yang disumbangkan oleh asid lemak tepu (SFA) daripada kumpulan lipid diasilgliserol (DAG) dan asid fosfatidik (PA). Aplikasi kemometrik didapati berjaya mendiskriminasikan lemak babi dari lemak-lemak lain setelah proses pemanasan dan biomarker FAs yang dikenalpasti secara tentatif. v ABSTRACT This research aims to evaluate the profile of lard using the chemometric method to discriminate lard from other types of fat after the heating-process and determination of biomarkers in lard. In this research, two types of samples were conducted. First, a total of 270 lard were collected from belly fats (BL), back fats (BK), and shoulder fats (SF) from northern, southern, and central regions of Peninsular Malaysia, which were measured using FTIR. PCA utilised the chemometrics evaluation on lard profiles using PCA and extended Hotelling T2. The scores vectors were found inside Hotelling T2 eclipses at first 3 PCs (99.5% confidence interval) and successfully validated by PCA projection. This indicates that the FTIR profile for lard is undifferentiated for all regions and body parts. Second, for discrimination purposes, a total of 60 fats, lard, chicken, beef, mutton, and plant have undergone twelve heating-process protocols (at 120 °C, 180 °C, 240 °C for 0.5. 1, 2 &3 hrs). In addition, each species without any heating was included in this study. After the heating-process, fat samples were analysed using FTIR, 1H-NMR, 13C-NMR, GC-FID, and LC-MS/MS. The classification was performed using multivariate classification (LDA, MDA, QDA & SVMDA) and multivariate regression (OSC-PLSR, PCR & PLSR) on FTIR and 1H-NMR. Multivariate classification found MDA, QDA, and SVMDA on 1H-NMR provided the best results for classifying lard from other fats. In the multivariate regression, the result of OSC-PLSR on FTIR (R2, adj. R2, RMSEC, RMSEV & MSEP; 0.985, 0.984, 0.049, 0.051 & 0.056) was found to outperform than the OSC-PLSR on 1H-NMR. Finally, profiling lard by 13C-NMR, GC- FID, and LC-MS/MS combined with the PCA were conducted to determine the biomarker of lard. The 13C-NMR-PCA data found that the lard differed against the chicken fats at C-2 of the TAG isomer, denoted by δ 34.21 and δ 62.10 resonances. Meanwhile, the GC-FID-PCA data found that animal fats and plant fats can differ after fatty acids )FAs( degradation of the cis into trans isomers. The lard heated at 180 °C (at 0.5, 1 & 2 hrs.) was found to differ significantly from others by LC-MS/MS-PCA, contributed by saturated fatty acids (SFA) of the diacylglycerol (DAG) and phosphatidic acids (PA) lipid classes. The application of chemometrics was found to discriminate lard against other fats after the heating-process successfully, and FAs were tentatively identified as biomarkers. iv الملخص يهدب هذا البحث إلى تقييم خصائص شحم الخنزير باستخدام طريقة القياس الكيميائي يميز شحم الخنزير نواع الأخرى من الدهون بعد عملية التسخين وتحديد المؤشرات الحيوية في شحم الخنزير. في هذا عن الأ ودهون الظهر )FS(شخم من ودهون الكتف 072البحث تم إجراء نوعين من العينات. أولا، المجموع يا، والتي كانت ) تم جمع من المناطق الشمالية والجنوبية والوسطى لشبه جزيرة ماليز LB(دهون البطن )KB( تم القياس با ستخدام تعريف RITF استخدمت. ACPتقييم القيا سات الكيميائية على ملفات تم العثور على ناقلات T gnilletoH2 .ACPو T gnilletoH2باستخدام الممتد شحم الخنزير ) وتم التحقق من صحته بنجاح 5.99قطع (فاصل ثقة ٪ ACP 3الدرجات داخل. يخسوف عند أول تنبؤ. يشير هذا إلى أن ملف تعريف وأجزاء RITFغير متمايز لجميع المناطق بواسطة لشحم الخنزير وم البقر والضأن والنبات دهون شحم دجاج، خضعت لح 06الجسم. ثانًيا، لأغراض التمييز، ما مجموعه .) srh 3 & 2 ,1 ,5.0عند C° 042 & C° 081 ,C° 021(لاثني عشر بروتوكوًلا لعملية التسخين بالإضافة إلى ذلك كان كل نوع دون أي تدفئة المدرجة في هذه الدراسة. بعد عملية التسخين، تم تحليل . ITFR ,1H-MNRCG ,- ,DIFCL-SM/SM, 31C-RMNعينات الدهون باستخدام ,ADM ,ADLتم التصنيف باستخدام تصنيف متعدد المتغيرات ( 1H-MNRو ITFR على ). أفضل RSLP & RCP ,RSLP-CSO). تم العثور على تصنيف (ADMVS & ADQ قدم الخنزير 1H-MNRشحم الخنزير على )C-1-ADQ & ADM ,ADMVS(النتائج لتصنيف jda. R2 589.0 = ( ITFR من الدهون الأخرى. في الانحدار متعدد المتغيرات، تكون نتيجة على 1H على كان ) R2 89.0 = ,4R ,940.0 = CESMR650.0 =PESM & 150.0=VESM جنًبا إلى ACPتفوق أداء لتحديد . أخيرًا، تم إجراء تنميط شحم الخنزير -RMN - RSLP-CSO العلامة الحيوية لدهن الخنزير. . أن شحم الخنزير31C-MNRCG ,-DIF ,=CLSM/SMجنب مع من GATُيشار إليها بصدى δ 01.26 12.43 ,يختلف عن 31C-MNR-ACP وجدت بيانا دهون الدجاج في أن الدهون الحيوانية والدهون. وفي الوقت نفسه، وجدت بيانات 2-Cأيزومر يمكن أن تختلف فيما بعد لرابطة الدول المستقلة إلى أيزومرات عابرة على , النباتية ACP-DIFCG تم العثور عليه بشكل ).srh 2 &,1 ,5.0(C° 081 SM/SM-CL. يسخن شحم الخنزير منsAF جلسرينمن دياسيل )AFS( أكثر وضوًحا للاختلاف بساهم عن طريق الأحماض الدهنية المشبعة الدهون. تم العثور على تطبيق القياسات الكيميائية للتمييز فئات )AP( وأحماض الفوسفاتيد )GAD( مؤقًتا تم تحديدها على أنها sAF بنجاح شحم ضد الدهون الأخرى بعد المرور بعملية التسخين، وكان .مؤشرات حيوية vii TABLE OF CONTENTS CONTENT PAGE AUTHOR DECLARATION ii ACKNOWLEDGEMENTS iii ABSTRAK iv ABSTRACT v صخلملا vi TABLE OF CONTENTS vii LIST OF TABLES xi LIST OF FIGURES xii LIST OF APPENDICES xiv LIST OF UNITS OF MEASUREMENTS xv LIST OF SYMBOLS xvi LIST OF EQUATIONS xvii LIST OF ABBREVIATIONS xviii CHAPTER 1: INTRODUCTION 1 1.1 Background of the Study 1 1.2 Statement of the Problem 4 1.3 Research Question 5 1.4 Scope of the Study 5 1.5 Objectives of the Study 7 1.6 The Significance of the Study 7 CHAPTER 2: LITERATURE REVIEW 9 2.1 Chemometrics Technique 9 2.1.1 Principal Component Analysis (PCA) 10 2.1.1.1 Hotelling’s T-Squared (T2) 12 2.1.1.2 PCA Projection 12 2.1.2 Multivariate Classification 14 2.1.2.1 Linear Discriminant Analysis (LDA) 15 2.1.2.2 Support Vector Machines (SVM) 17 2.1.2.3 Evaluation of Multivariate Classification 18 2.1.3 Multivariate Regression 19 2.1.3.1 Principal Component Regression (PCR) 20 2.1.3.2 Partial Least Squares (PLS) 21 2.1.3.3 Orthogonal Signal Correction Partial Least Squares (OSC-PLS) 23 2.1.3.4 Multivariate Regression Coefficient as Important Features Assessment 23 2.1.3.5 Evaluation of Multivariate Regression Analysis 24 2.1.4 Data Pre-processing 27 2.2 Application of Chemometrics in Lard Profiles for Halal Authentication 29 2.2.1 FTIR spectroscopy in Authenticity Studies 31 2.2.2 NMR (1H & 13C) 39 2.2.3 Chromatography (GC-FID, GC-MS, & HPLC) 41 viii CHAPTER 3: METHODOLOGY 45 3.1 Experimental Design 45 3.2 Profiling of Lard from Collected Pig Samples at Northern, Central, and Southern Malaysia 47 3.2.1 Chemicals and Materials 47 3.2.1.1 Chemicals 47 3.2.1.2 Lard Samples 47 3.2.2 Instruments 48 3.2.3 Samples Preparation 48 3.2.3.1 Preparation of Lard Samples 48 3.2.3.2 Extraction of Lard Samples 49 3.2.3.3 Samples Preparation for FTIR Analysis 49 3.3 Profiling on Pork (Lard), Chicken, Beef, Mutton, and Plant Fats after Heating-Process 51 3.3.1 Chemicals and Materials 51 3.3.1.1 Chemicals 51 3.3.1.2 FAMEs Standard 51 3.3.2 Samples of Lard and Selected Fats 52 3.3.3 Apparatus and Instruments 52 3.3.4 Sample Preparation of Animals Fats 54 3.3.4.1 Heating-Process Protocols 54 3.3.4.2 Extraction of Fats after Heating-Process 55 3.3.4.3 Sample Preparation of FTIR Analysis 55 3.3.4.4 Sample Preparation of 1H-NMR Analysis 56 3.3.4.5 Sample Preparation of 13C-NMR Analysis 56 3.3.4.6 Sample Preparation of GC-FID Analysis (FAs Methylation) 56 3.3.4.7 Sample Preparation of LC-MS/MS Analysis 57 3.3.5 Fats Analysis 58 3.3.5.1 FTIR Analysis 58 3.3.5.2 NMR (1H & 13C) Analysis 58 3.3.5.3 GC-FID Analysis 58 3.3.5.4 LC-MS/MS Analysis 59 3.4 Data Analysis 60 3.4.1 Data Pre-processing 60 3.4.1.1 Data of FTIR 61 3.4.1.2 Data of 1H-NMR. 61 3.4.1.3 Data of 13C-NMR 61 3.4.1.4 Data of GC-FID 62 3.4.1.5 Data of LC-MS/MS 62 3.4.2 Chemometrics Techniques 62 3.4.2.1 Kennard-Stone (K-S) Algorithm Selection 64 3.4.2.2 Flowchart of Chemometrics Evaluation 65 CHAPTER 4: RESULTS AND DISCUSSION 62 4.1 Profiling of Lard from Collected Pig Samples at Northern, Central, and Southern Malaysia using FTIR combined with Principal Component Analysis (PCA) 62 4.1.1 Introduction 62 ix 4.1.2 Data Pre-processing 69 4.1.3 Principal Component Analysis (PCA) 72 4.1.4 Hotelling T2 Similarity Assessment of MSC-PCA Models 76 4.1.5 Prediction of Lard by MSC-PCA Projection 78 4.2 Discrimination of Pigs (Lard), Chicken, Beef, Mutton, and Plant Fats After Heating-Process using FTIR, 1H-NMR, and 13C-NMR with Chemometrics Techniques 84 4.2.1 Introduction 84 4.2.2 Profiling of Lard and Selected Edible Fats After the Heating- Process. 85 4.2.2.1 FTIR 85 4.2.2.2 1H-NMR 87 4.2.3 Data Pre-processing 89 4.2.3.1 FTIR 89 4.2.3.2 1H-NMR 91 4.2.4 PCA 92 4.2.4.1 PCA of FTIR data 92 4.2.4.2 PCA of 1H-NMR data 97 4.2.5 Multivariate Classification (MVC) 99 4.2.5.1 Class Sensitivity 101 4.2.5.2 Class Specificity 102 4.2.5.3 Classification Measures of the MCC. 104 4.2.6 Multivariate Regression (MVR) 105 4.2.6.1 FTIR 105 4.2.6.2 1H-NMR 107 4.2.6.3 Comparison between FTIR and 1H-NMR Multivariate Regression 108 4.2.7 Correlation Chemical Features of Lard after Heating-Process by OSC-PLSR 110 4.2.7.1 FTIR 110 4.2.7.2 1H-NMR 111 4.2.8 Profiling of 13C-NMR Lard vs Chicken fats 114 4.2.9 Data Pre-processing of 13C-NMR Spectra 115 4.2.10 PCA of 13C-NMR Lard vs Chicken fats 117 4.3 Evaluation on Fatty Acids (FAs) of Lard and Selected Fats after Heating-Process using GC-FID and LC-MS/MS combined with Principal Component Analysis (PCA) 121 4.3.1 Gas Chromatography with Flame Ionization Detection (GC- FID) 121 4.3.1.1 Introduction 121 4.3.1.2 Fatty Acids (FAs) Profiling 121 4.3.1.3 PCA of GC-FID 123 4.3.2 Liquid Chromatography with Tandem Mass Spectrometry (LC- MS/MS) 128 4.3.2.1 Introduction 128 4.3.2.2 Profiling of Lard and Selected Fats by LC-MS/MS 128 4.3.2.3 Glycerolipids of the PCA by LC-MS/MS (LC- MS/MS-GL-PCA) 129 x 4.3.2.4 Glycerophospholipids of the PCA by LC-MS/MS (LC-MS/MS-GPPL-PCA) 133 CHAPTER 5: CONCLUSION AND RECOMMENDATION 137 5.1 Profiling of Lard from Collected Pig Samples at Northern, Central, and Southern in Malaysia using FTIR combined with Principal Component Analysis (PCA) 137 5.2 Discrimination of Pigs (Lard), Chicken, Beef, Mutton, and Plant Fats after Heating-Process using FTIR, 1H-NMR, and 13C-NMR with Chemometrics Techniques 138 5.3 Evaluation on Fatty Acids (FAs) of Lard and Selected Fats after Heating-Process using GC-FID and LC-MS/MS combined with Principal Component Analysis (PCA) 140 REFERENCES 142 APPENDICES 163 xi LIST OF TABLES Tables Page Table 2.1: Application Chemometrics and FTIR on Profiles Lard. 38 Table 3.1: Chemicals. 47 Table 3.2: Total Pigs Fat of the Body Parts and Regions. 48 Table 3.3: Chemicals. 51 Table 3.4: Heating-Process Protocols. 55 Table 4.1: Comparison between Four Different Significance Levels (α). 77 Table 4.2: Multivariate Classification on FTIR and 1H-NMR. 100 Table 4.3: Statistics for Multivariate Regression. 108 Table 4.4: Wavenumbers Contributed to the Lard. 111 Table 4.5: Summary of the δ by 1H-NMR-OCR-PLS. 113 Table 4.6: X-correlation Loadings Plot of the GC-FID-PCA. 125 Table 4.7: X-correlation Loadings Plot of All Fats (LC-MS/MS-GL-PCA). 130 Table 4.8: X-correlation Loadings Plot (LC-MS/MS-GL-PCA). 132 Table 4.9: X-correlation Loadings Plot (LC-MS/MS-GPPL-PCA). 134 xii LIST OF FIGURES Figures Page Figure 2.1: Illustration of Boundaries LDA (left) vs QDA (right). 16 Figure 2.2: Basic Idea of the Kernel Function in SVM. 17 Figure 3.1: Framework of Lard Profiling. 46 Figure 3.2: Chemometrics Technique for Lard Profiles. 66 Figure 3.3: Chemometrics Technique for the Heated Lard Profiles. 67 Figure 4.1: Comparison of the Untreated and Pre-processing FTIR Spectra. 69 Figure 4.2: Comparison of the FTIR-PCA Before and After Pre-processing. 72 Figure 4.3: Determination of Outliers by MSC-PCA. 76 Figure 4.4: Projection of the Test Set into MSC-PCA. 79 Figure 4.5: Influence Graph Plots F-Residual vs Hoteling T2. 80 Figure 4.6: Total Residual Variance of MSC-PCA. 81 Figure 4.7: Comparison of FTIR Spectra. 85 Figure 4.8: The Basic Structure of TAG. 87 Figure 4.9: 1H-NMR Spectra. 88 Figure 4.10: Comparison of Normalised vs 2nd DSG FTIR Spectra. 89 Figure 4.11: 1H-NMR Spectra After Pre-processing. 91 Figure 4.12: The Normalised-PCA Overview of the FTIR Data. 93 Figure 4.13: The 2nd DSG-PCA Overview of the FTIR Data. 94 Figure 4.14: The 1H-NMR-PCA of Untreated and After Pre-processing. 97 Figure 4.15: Multivariate Regression on FTIR data. 105 Figure 4.16: Multivariate Regression on 1H-NMR. 107 Figure 4.17: Variables Selection by p-Correlation. 112 Figure 4.18: Stacked individual 13C-NMR Spectra of Lard vs Chicken Fats. 115 Figure 4.19: 13C-NMR Spectra of the Lard & Chicken Fats. 116 xiii Figure 4.20: 13C-NMR Spectra After Baseline Pre-Processing. 117 Figure 4.21: 13C-NMR-PCA of the Lard vs Chicken Fats. 118 Figure 4.22: Heatmap of 13C-NMR-PCA. 119 Figure 4.23: TAG Isomers. 120 Figure 4.24: Box Plot of FAs 121 Figure 4.25: GC-FID-PCA. 124 Figure 4.26: Bar Plot Sum of Heat Values. 126 Figure 4.27: The Profiles of Fats by LC-MS/MS. 129 Figure 4.28: GL-PCA on LC-MS/MS Data of All Fats. 129 Figure 4.29: GL-PCA on LC-MS/MS of Lard vs Chicken. 131 Figure 4.30: GPPL-PCA on LC-MS/MS Data of All Fats. 133 xiv LIST OF APPENDICES Appendices Page Appendix 1: PCA Overview of Untreated FTIR. 163 Appendix 2: Loading Plots of FTIR-PCA on Fats. 164 Appendix 3: Important Variables Selection of FTIR-OSC-PLSR. 165 Appendix 4: Assignment of the Main Resonances in the 1H- NMR. 166 Appendix 5: Important Variables Selection of 1H-NMR OSC-PLSR. 167 Appendix 6: Assignment of the Main Resonances in the 13C-NMR. 168 Appendix 7: Estimation of the 13C-NMR. 169 Appendix 8: Standard of FAME Integration. 170 Appendix 9: Typical Chromatogram of Lard. 171 Appendix 10: Most Prominent FAs GC-FID. 172 Appendix 11: Scores Plot of GC-FID-PCA and Clusters. 173 Appendix 12: Lipid Classes of the LC-MS/MS. 174 Appendix 13: Typical Spectra of Glycerolipids (GL). 176 Appendix 14: Typical Spectra of Glycerophospholipids (GPPL). 177 Appendix 15: List of Publications. 179 xv LIST OF UNITS OF MEASUREMENTS °C degree Celsius g gram Hertz Hz kg kilogram L litre mg milligram mL millilitre M molar mol mole ppm parts per million µL microlitre . xvi LIST OF SYMBOLS C hyperplane SVM parameter df degree of freedom hr hour hrs hours min minutes sec second vs versus α significance level δ chemical shift ε epsilon υ Nu xvii LIST OF EQUATIONS Equations Page 2.1 10 2.2 13 2.3 14 2.4 16 2.5 17 2.6 18 2.7 18 2.8 19 2.9 19 2.10 20 2.11 21 2.12 21 2.13 21 2.14 22 2.15 22 2.16 22 2.17 23 2.18 24 2.19 24 2.20 25 3.1 64 3.2 64 xviii LIST OF ABBREVIATIONS 1D One-Dimensional 2D Two-Dimensional 3D Three-Dimensional ANN Artificial Neural Network. ANOVA Analysis of Variance ATR Attenuated Total Reflectance CI Confidence Interval COW Correlation Optimized Warping CRM Certified Reference Material DA Discriminant Analysis DAG Diacylglycerol DSG Derivatives Savitzky-Golay EA/IRMS Elemental Analyzer/Isotope Ratio Mass Spectrometry EVCO Extra Virgin Coconut Oil FAME Fatty Acid Methyl Ester FAMEs Fatty Acid Methyl Esters (plural) FA Fatty Acid FAs Fatty Acids (plural) FTIR Fourier Transform Infrared GC Gas Chromatography GC × GC-TOF-MS Dimensional Gas Chromatography Coupled with Time-Of-Flight Mass Spectrometry GC-FID Gas Chromatography with Flame Ionization Detection GC-MS Gas Chromatography-Mass Spectrometry GL Glycerolipid GLC Gas Liquid Chromatography GPPL Glycerophospholipid HCA Hierarchical Cluster Analysis HPLC High-Performance Liquid Chromatography IPA Isopropyl Alcohol IR Infrared k-NN k-Nearest Neighbours K-S Kennard-Stone LC-MS/MS Liquid Chromatography with Tandem Mass Spectrometry LDA Linear Discriminant Analysis MAG Monoacylglycerol MDA Mahalanobis Discriminant Analysis MSC Multiplicative Scatter Correction MSEP Mean Standard Error Prediction MUFA Monounsaturated Fatty Acids xix MVR Multivariate Regression NIR Near-Infrared Spectroscopic NMR Nucleus Magnetic Resonance OPLS Orthogonal Partial Least Squares OPLS-DA Orthogonal Partial Least Squares Discriminant Analysis OPLSR Orthogonal Partial Least Squares Regression OSC-PLSR Orthogonal Signal Correction Partial Least Squares Regression PA Phosphatidic Acids PC Principal Component PCs Principal Components (plural) PCA Principal Component Analysis PCh Phosphatidylcholines PCR Principal Component Regression PG Phosphatidyl-Glycerol PI Phosphatidylinositol PLS Partial Least Squares PLSR Partial Least Squares Regression PUFA Polyunsaturated Fatty Acids QDA Quadratic Discriminant Analysis R2 Coefficient of Determination RBF Radial Basis Function RF Random Forest RMSEC Root Mean Square Error of Calibration RMSEV Root Mean Square Error of Validation SFA Saturated Fatty Acids SVM Support Vector Machines SVMDA Support Vector Machines Discriminant Analysis SVMR Support Vector Machines Regression TAG Triacylglycerol TAGs Triacylglycerols (plural) TLC Thin Layer Chromatography TMS Tetramethylsilane VCO Virgin Coconut Oil