RESULTS: Comparison of the PLS and RF showed that RF exhibited poorer generalization and hence poorer predictive performance. Both the regression coefficient of PLS and the variable importance of RF revealed that quercetin and kaempferol derivatives, caffeic acid and vitexin-2-O-rhamnoside were significant towards the tested bioactivities. Furthermore, principal component analysis (PCA) and partial least squares-discriminant analysis (PLS-DA) results showed that sonication and absolute ethanol are the preferable extraction method and ethanol ratio, respectively, to produce N. oleracea extracts with high phenolic levels and therefore high DPPH scavenging and α-glucosidase inhibitory activities.

OBJECTIVE: Evaluate the relationship between the chemical composition of C. nutans and its anti-inflammatory properties using nuclear magnetic resonance (NMR) metabolomics approach.

METHODOLOGY: The anti-inflammatory effect of C. nutans air-dried leaves extracted using five different binary extraction solvent ratio and two extraction methods was determined based on their nitric oxide (NO) inhibition effect in lipopolysaccharide-interferon-gamma (LPS-IFN-γ) activated RAW 264.7 macrophages. The relationship between extract bioactivity and metabolite profiles and quantifications were established using 1 H-NMR metabolomics and liquid chromatography-tandem mass spectrometry (LC-MS/MS). The possible metabolite biosynthesis pathway was constructed to further strengthen the findings.

RESULTS: Water and sonication prepared air-dried leaves possessed the highest NO inhibition activity (IC50  = 190.43 ± 12.26 μg/mL, P least square (PLS) biplot suggested that sulphur containing glucoside, sulphur containing compounds, phytosterols, triterpenoids, flavones and some organic and amino acids were among the potential NO inhibitors. LC-MS/MS targeted quantification further supported sonicated water extract was among the extract that possessed the most abundant C-glycosyl flavones.

OBJECTIVE: To determine and quantify lard as an adulterant in a binary blend with palm oil in a cosmetic soap formulations by FT-IR and multivariate analysis.

METHODS: Fatty acids in lard, palm oil and binary blends were extracted via liquid-liquid extraction and were subjected to FTIR spectrometry, combined with principal component analysis (PCA) and discriminant analysis (DA) for the classification of lard in cosmetic soap formulations via two DA models: Model A (percentage of lard in cosmetic soap) and Model B (porcine and non-porcine cosmetic soap). Linear regression (MLR), partial least square regression (PLS-R) and principal components regression (PCR) were used to assess the degree of adulteration of lard in the cosmetic soap.

FINDINGS: The FTIR spectrum of palm oil slightly differed from that of lard at the wavenumber range of 1453 cm -1 and 1415 cm -1 in palm oil and lard, respectively, indicating the bending vibrations of CH2 and CH3 aliphatic groups and OH carboxyl group respectively. Both of the DA models could accurately classify 100% of cosmetic soap formulations. Nevertheless, less than 100% of verification value was obtained when it was further used to predict the unknown cosmetic soap sample suspected of containing lard or a different percentage of lard. The PCA for Model A and Model B explained a similar cumulative variability (CV) of 92.86% for the whole dataset. MLR and PCR showed the highest determination coefficient (R2) of 0.996, and the lowest relative standard error (RSE) and mean square error (MSE), indicating that both regression models were effective in quantifying the lard adulterant in cosmetic soap.

METHODS: We use Non-linear Iterative Partial Least Squares to perform the data dimensionality reduction, Self-Organizing Map technique for clustering task and ensembles of Neuro-Fuzzy Inference System for predicting the hepatitis disease. We also use decision trees for the selection of most important features in the experimental dataset. We test our method on a real-world dataset and present our results in comparison with the latest results of previous studies.

RESULTS: The results of our analyses on the dataset demonstrated that our method performance is superior to the Neural Network, ANFIS, K-Nearest Neighbors and Support Vector Machine.