Comparison of the Performance of Partial Least Squares and Support Vector Regressions for Predicting Fatty Acids and Fatty Acid Classes in Marine Oil Dietary Supplements by Using Vibrational Spectroscopic Data.

ABSTRACT: Simple, fast, and accurate analytical techniques for verifying the accuracy of label declarations for marine oil dietary supplements containing eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA) are required because of the increased consumption of these products. We recently developed broad-based partial least squares regression (PLS-R) models to quantify six fatty acids (FAs) and FA classes by using the spectroscopic data from a portable Fourier transform infrared (FTIR) device and a benchtop Fourier transform near infrared (FT-NIR) spectrometer. We developed an improved quantification method for these FAs and FA classes by incorporating a nonlinear calibration approach based on the machine learning technique support vector machines. For the two spectroscopic methods, high accuracy in prediction was indicated by low root mean square error of prediction and by correlation coefficients (R2) close to 1, indicating excellent model performance. The percent accuracy of the support vector regression (SV-R) model predicted values for EPA and DHA in the reference material was 90 to 110%. In comparison to PLS-R, SV-R accuracy for prediction of FA and FA class concentrations was up to 2.4 times higher for both ATR-FTIR and FT-NIR spectroscopic data. The SV-R models also provided closer agreement with the certified and reference values for the prediction of EPA and DHA in the reference standard. Based on our findings, the SV-R methods had superior accuracy and predictive quality for predicting the FA concentrations in marine oil dietary supplements. The combination of SV-R with ATR-FTIR and/or FT-NIR spectroscopic data can potentially be applied for the rapid screening of marine oil products to verify the accuracy of label declarations.

Rapid Prediction of Low (<1%) trans Fat Content in Edible Oils and Fast Food Lipid Extracts by Infrared Spectroscopy and Partial Least Squares Regression.

The United States Food and Drug Administration (FDA) ruled that partially hydrogenated oils (PHO), the major dietary source of industrially produced trans fat (TF), were no longer "generally recognized as safe (GRAS)" for any use in human food. Consequently, the objective of this study was to develop a rapid screening procedure using attenuated total reflection Fourier-transform infrared (ATR-FTIR) spectroscopy in conjunction with partial least squares regression (PLSR) for the quantitative and accurate prediction of low concentrations of trans fatty acid (TFAs) (<1% of total fatty acids (FAs)). Broad-based calibration models were developed for a combined set of samples consisting of edible oils and fast food lipid extracts. Predicted concentrations of TFAs in the two matrices showed good correlation with the primary reference data generated by gas chromatography (GC) (R2 > 0.99) and high accuracy as evidenced by low root-mean-square error of cross-validation (RMSECV) values. The lowest TFA concentration, determined by GC to be 0.13% of total FAs, was accurately predicted by ATR-FTIR/PLSR as 0.18% of total FAs. This simple, rapid ATR-FTIR/PLSR methodology has the potential for use as a screening alternative to conventional gas chromatographic methods for predicting the TFA content of edible oils and food lipid extracts for regulatory purposes and quality control of raw material and processed food. PRACTICAL APPLICATIONS: FDA ruled that partially hydrogenated oils were no longer "generally recognized as safe (GRAS)" for any use in human food. Consequently, we have proposed a rapid screening procedure, based on infrared spectroscopy and chemometrics, to rapidly and accurately predict low concentrations of trans fatty acids (<1% of total fatty acids) in edible oils and food lipid extracts.

First Application of Newly Developed FT-NIR Spectroscopic Methodology to Predict Authenticity of Extra Virgin Olive Oil Retail Products in the USA.

Economically motivated adulteration (EMA) of extra virgin olive oils (EVOO) has been a worldwide problem and a concern for government regulators for a long time. The US Food and Drug Administration (FDA) is mandated to protect the US public against intentional adulteration of foods and has jurisdiction over deceptive label declarations. To detect EMA of olive oil and address food safety vulnerabilities, we used a previously developed rapid screening methodology to authenticate EVOO. For the first time, a recently developed FT-NIR spectroscopic methodology in conjunction with partial least squares analysis was applied to commercial products labeled EVOO purchased in College Park, MD, USA to rapidly predict whether they are authentic, potentially mixed with refined olive oil (RO) or other vegetable oil(s), or are of lower quality. Of the 88 commercial products labeled EVOO that were assessed according to published specified ranges, 33 (37.5%) satisfied the three published FT-NIR requirements identified for authentic EVOO products which included the purity test. This test was based on limits established for the contents of three potential adulterants, oils high in linoleic acid (OH-LNA), oils high in oleic acid (OH-OLA), palm olein (PO), and/or RO. The remaining 55 samples (62.5%) did not meet one or more of the criteria established for authentic EVOO. The breakdown of the 55 products was EVOO potentially mixed with OH-LNA (25.5%), OH-OLA (10.9%), PO (5.4%), RO (25.5%), or a combination of any of these four (32.7%). If assessments had been based strictly on whether the fatty acid composition was within the established ranges set by the International Olive Council (IOC), less than 10% would have been identified as non-EVOO. These findings are significant not only because they were consistent with previously published data based on the results of two sensory panels that were accredited by IOC but more importantly each measurement/analysis was accomplished in less than 5 min.

Rapid Prediction of Fatty Acid Content in Marine Oil Omega-3 Dietary Supplements Using a Portable Fourier Transform Infrared (FTIR) Device and Partial Least-Squares Regression (PLSR) Analysis.

Using a portable field device, a Fourier transform infrared spectroscopy (FTIR) and partial least-squares regression (PLSR) method was developed for the rapid (<5 min) prediction of major and minor fatty acid (FA) concentrations in marine oil omega-3 dietary supplements. Calibration models were developed with 174 gravimetrically prepared samples. These models were tested using an independent validation set of dietary supplements. FAs analyzed included eicosapentaenoic acid (EPA) and docosahexaenoic acid (DHA); the sums of saturated, branched-chain, and monounsaturated FAs; and n-6, n-4, n-3, n-1, and trans polyunsaturated FA. The spectral ranges 650-1500 or 650-1500 and 2800-3050 cm-1 provided reliable predictions for FA components in 34 neat oil products: standard error of prediction, 0.73-1.58%; residual predictive deviation, 6.41-12.6. This simple, nondestructive quantitative method is a rapid screening tool and a time and cost-saving alternative to gas chromatography for verifying label declarations and in quality control.

Microarray chip development using infrared imaging for the identification of catfish species.

Several families of catfish species are extensively aquacultured around the world; however, only those from the family Ictaluridae can be labeled as catfish in the United States. Non-Ictalurid catfish species that are marketed as "catfish" in the USA are considered misbranded. Misbranding in general has led to an increased interest in developing deoxyribonucleic acid (DNA)-based methods such as DNA barcoding, polymerase chain reaction restriction fragment length polymorphism, and DNA microarrays with fluorescence detection for the identification of fish species. In this proof-of-concept study, DNA microarrays coupled with a newly developed mid-infrared imaging detection method were applied to the identification of seven species of catfish for the first time. Species-specific DNA probes targeting three regions per species of the cytochrome c oxidase 1 (barcoding) gene were developed and printed as microarrays on glass slides. Deoxyribonucleic acid targets labeled with biotin were hybridized to their complementary probes using a strategy that allowed the selective formation of a silver layer on hybridized spots needed for detection. Using this three-probe format, the seven species were all identified correctly, even when a limited number of false positive spots were observed. Raman spectroscopy was employed to further characterize the arrays.

Profile of trans fatty acids (FAs) including trans polyunsaturated FAs in representative fast food samples.

The content of trans fat in foods is most commonly determined by summing the levels of individual trans fatty acids (FAs), analyzed as FA methyl esters (FAME) by gas chromatography. Current Official Methods of the American Oil Chemists' Society (AOCS) enable quantitation of total trans fat in foods but were not designed for the determination of transFA isomeric compositions. In the present study, the content of trans fat in 32 representative fast food samples ranged from 0.1 to 3.1 g per serving, as determined according to AOCS Official Method Ce 1j-07. Further analysis of FAME using the 200 m SLB-IL111 ionic liquid column yielded quantitative results of total, trans, saturated, and cis unsaturated fat that were comparable to those of Method Ce 1j-07 and also allowed for the complementary determination of individual trans 18:1, trans 18:2, and trans 18:3 FA isomeric compositions under conditions suitable for routine sample analysis.

Evaluation of highly polar ionic liquid gas chromatographic column for the determination of the fatty acids in milk fat.

The SLB-IL111, a new ionic liquid capillary column for gas chromatography available from Supelco Inc., was recently shown to provide enhanced separation of unsaturated geometric and positional isomers of fatty acid (FAs) when it was compared to cyanopropylsiloxane (CPS) columns currently recommended for the analysis of fatty acid methyl esters (FAMEs). A 200 m SLB-IL111 capillary column, operated under a combined temperature and eluent flow gradient, was successfully used to resolve most of the FAs contained in milk fat in a single 80 min chromatographic separation. The selected chromatographic conditions provided a balanced, simultaneous separation of short-chain (from 4:0), long-chain polyunsaturated fatty acids (PUFAs), and most of the unsaturated FA positional/geometric isomers contained in milk fat. Among the monounsaturated fatty acids (MUFAs), these conditions separated t11-18:1 and t10-18:1 FAs, the two most abundant trans fatty acids (t-FA) contained in most dairy products. These t-FAs reportedly have different biological activities. The conjugated linoleic acid (CLA) isomers commonly found in dairy products were separated from each other, including t7,c9-18:2 from c9,t11-18:2, which eliminated the need for their complementary silver ion HPLC analysis. The application of the SLB-IL111 column provided a complementary elution profile of FAMEs to those obtained by CPS columns, allowing for a more comprehensive FA analysis of total milk fat. The FAMEs were identified by the use of available reference materials, previously synthesized and characterized reference mixtures, and prior separations of the milk fat FAMEs by silver ion chromatography based on the number/geometry of double bonds.

Separation characteristics of fatty acid methyl esters using SLB-IL111, a new ionic liquid coated capillary gas chromatographic column.

The ionic liquid SLB-IL111 column, available from Supelco Inc., is a novel fused capillary gas chromatography (GC) column capable of providing enhanced separations of fatty acid methyl esters (FAMEs) compared to the highly polar cyanopropyl siloxane columns currently recommended for the separation of cis- and trans isomers of fatty acids (FAs), and marketed as SP-2560 and CP-Sil 88. The SLB-IL111 column was operated isothermal at 168°C, with hydrogen as carrier gas at 1.0 mL/min, and the elution profile was characterized using authentic GC standards and synthetic mono-unsaturated fatty acids (MUFAs) and conjugated linoleic acid (CLA) isomers as test mixtures. The SLB-IL111 column provided an improved separation of cis- and trans-18:1 and cis/trans CLA isomers. This is the first direct GC separation of c9,t11- from t7,c9-CLA, and t15-18:1 from c9-18:1, both of which previously required complimentary techniques for their analysis using cyanopropyl siloxane columns. The SLB-IL111 column also provided partial resolution of t13/t14-18:1, c8- from c6/c7-18:1, and for several t,t-CLA isomer pairs. This column also provided elution profiles of the geometric and positional isomers of the 16:1, 20:1 and 18:3 FAMEs that were complementary to those obtained using the cyanopropyl siloxane columns. However, on the SLB-IL111 column the saturated FAs eluted between the cis- and trans MUFAs unlike cyanopropyl siloxane columns that gave a clear separation of most saturated FAs. These differences in elution pattern can be exploited to obtain a more complete analysis of complex lipid mixtures present in ruminant fats.