Principal component analysis and hierarchical cluster analysis for homogeneity evaluation during the preparation of a wheat flour laboratory reference material for inorganic analysis

Preparation and characterization of new reference material for inorganic analysis of pumpkin seed flour - An interlaboratory program

This work describes preparing a certified reference material for pumpkin seed flour for use in inorganic analyses of vegetable foods. The elements studied were potassium, magnesium, phosphorus, zinc, copper, iron, manganese, and calcium. The ISO Guides were employed in the homogeneity and stability tests, interlaboratory program, and assignment of uncertainty value. The material was stable for 12 months at temperatures of -10 °C, 25 °C, and 45 °C, with minor uncertainties. Chemometric techniques, such as principal component analysis (PCA) and hierarchical clustering analysis (HCA), were employed in the homogeneity and stability studies. The confidence ellipse technique allowed a simple and rapid evaluation of the data furnished by the collaborating laboratories regarding the precision of the results. The findings showed that pumpkin seed flour could be considered a raw material for preparing reference material, considering its availability and characteristics of its matrix, including the presence of trace and essential elements.

DWSSA: Alleviating over-smoothness for deep Graph Neural Networks

Graph Neural Networks (GNNs) have demonstrated great potential in achieving outstanding performance in various graph-related tasks, e.g., graph classification and link prediction. However, most of them suffer from the following issue: shallow networks capture very limited knowledge. Prior works design deep GNNs with more layers to solve the issue, which however introduces a new challenge, i.e., the infamous over-smoothness. Graph representation over emphasizes node features but only considers the static graph structure with a uniform weight are the key reasons for the over-smoothness issue. To alleviate the issue, this paper proposes a Dynamic Weighting Strategy (DWS) for addressing over-smoothness. We first employ Fuzzy C-Means (FCM) to cluster all nodes into several groups and get each node's fuzzy assignment, based on which a novel metric function is devised for dynamically adjusting the aggregation weights. This dynamic weighting strategy not only enables the intra-cluster interactions, but also inter-cluster aggregations, which well addresses undifferentiated aggregation caused by uniform weights. Based on DWS, we further design a Structure Augmentation (SA) step for addressing the issue of underutilizing the graph structure, where some potentially meaningful connections (i.e., edges) are added to the original graph structure via a parallelable KNN algorithm. In general, the optimized Dynamic Weighting Strategy with Structure Augmentation (DWSSA) alleviates over-smoothness by reducing noisy aggregations and utilizing topological knowledge. Extensive experiments on eleven homophilous or heterophilous graph benchmarks demonstrate the effectiveness of our proposed method DWSSA in alleviating over-smoothness and enhancing deep GNNs performance.

Comprehensive Evaluation of Nutritional Qualities of Chinese Cabbage (Brassica rapa ssp. pekinensis) Varieties Based on Multivariate Statistical Analysis

In order to make the identification and utilization of nutritional quality components in Chinese cabbage more predictive, to obtain ideal raw materials, and to help screen functional Chinese cabbage varieties that have high nutritional value, we conducted a comprehensive evaluation of the nutritional quality of different Chinese cabbage varieties. In this study, 17 nutritional quality indexes of 35 Chinese cabbage varieties, including crude fiber (CF), crude protein (CP), vitamin C (VC), glucose (Glc), fructose (Fru), malic acid (MA), citric acid (CA), oxalic acid (OA), total amino acid (TAA) and CA, K, Mg, P, Cu, Fe, Mn and Zn, were analyzed using diversity analysis, correlation analysis, principal component analysis, membership function analysis and cluster analysis. The results showed that there were different degrees of variation in the 17 nutritional quality indexes, and the coefficients of variation ranged from 11.45% to 91.47%. The correlation analysis found that there were significant or extremely significant correlations between different nutrient elements of Chinese cabbage, which indicated that principal component analysis could be carried out, and the comprehensive score (D value) of different materials could be obtained using principal component analysis and the membership function method. The nutritional quality of Chinese cabbage was classified into five grades by cluster analysis. Finally, a mathematical model for evaluating the nutritional quality of Chinese cabbage was established using the D value and multiple stepwise regression methods, and 10 key indexes were selected from the 17 indexes, which could be used for the rapid identification of the nutritional quality of Chinese cabbage. This study provided a theoretical basis for the nutritional quality evaluation and variety breeding of Chinese cabbage.

ATR-FTIR combined with machine learning for the fast non-targeted screening of new psychoactive substances

Due to the diversity and fast evolution of new psychoactive substances (NPS), both public health and safety are threatened around the world. Attenuated total reflection-Fourier transform infrared spectroscopy (ATR-FTIR), which serves as a simple and rapid technique for targeted NPS screening, is challenging with the rapid structural modifications of NPS. To achieve the fast non-targeted screening of NPS, six machine learning (ML) models were constructed to classify eight categories of NPS, including synthetic cannabinoids, synthetic cathinones, phenethylamines, fentanyl analogues, tryptamines, phencyclidine types, benzodiazepines, and "other substances" based on the 1099 IR spectra data items of 362 types of NPS collected by one desktop ATR-FTIR and two portable FTIR spectrometers. All these six ML classification models, including k-nearest neighbor (KNN), support vector machine (SVM), random forest (RF), extra trees (ET), voting, and artificial neural networks (ANNs) were trained through cross validation, and f1-scores of 0.87-1.00 were achieved. In addition, hierarchical cluster analysis (HCA) was performed on 100 synthetic cannabinoids with the most complex structural variation to investigate the structure-spectral property relationship, which leads to a summary of eight synthetic cannabinoid sub-categories with different "linked groups". ML models were also constructed to classify eight synthetic cannabinoid sub-categories. For the first time, this study developed six ML models, which were suitable for both desktop and portable spectrometers, to classify eight categories of NPS and eight synthetic cannabinoids sub-categories. These models can be applied for the fast, accurate, cost-effective, and on-site non-targeted screening of newly emerging NPS with no reference data available.

The use of ANOVA-PCA and DD-SIMCA in the development of corn flour laboratory reference materials

The development of a collaborative study as a requirement for the preparation of a laboratory reference material candidate is reported in this paper. The evaluation was performed by 13 laboratories invited to quantify the calcium, potassium, magnesium, phosphorus, copper, iron, manganese and zinc; 8 of them presented results for all the analytes under investigation. The data were statistically analyzed by applying the z-score robust technique as recommended by ISO Guide 35. For the potassium element, laboratories 4 and 13 presented questionable results. Laboratory 5 proved to be unsatisfactory for calcium and zinc. ANOVA-PCA and DD-SIMCA were also applied to evaluate stability and interlaboratory studies results, respectively. It has been demonstrated that multivariate data analysis can be successfully applied as an alternative method to the recommendations made by ISO 13528 and ISO Guide 35 with defined confidence intervals.

Quality assessment of Moringa seed shells based on fingerprinting using HPLC-DAD

A fingerprint analysis method was established for the quality control of Moringa seed shells by high-performance liquid chromatography with diode array detection (HPLC–DAD). The HPLC–DAD separation was performed on a Thermo Hypersil Gold C18 (4.6 mm × 250 mm, 5 μm) column by gradient elution with acetonitrile–water as mobile phase. The fingerprint of Moringa seed shells was established with good precision, reproducibility, and stability obtaining within 60 min, and 13 common peaks in the fingerprint were designed. Similarity analysis, principal component analysis (PCA), and hierarchical clustering analysis (HCA) were carried out to analyze the obtained fingerprints. The similarity among 11 batches of samples in addition to No. 5 and 6 was no less than 0.92. Eleven samples could be classified into 2 clusters. The HPLC fingerprint technology and application of chemical pattern recognition can provide a more comprehensive reference for the quality control of medicinal plants.

Quality Assessment of Ojeok-San, a Traditional Herbal Formula, Using High-Performance Liquid Chromatography Combined with Chemometric Analysis

Ojeok-san (OJS) is a traditional herbal formula consisting of 17 herbal medicines that has been used to treat various disorders. In this study, quantitative analytical methods were developed using high-performance liquid chromatography equipped with a photodiode array detector to determine 19 marker compounds in OJS preparations, which was then combined with chemometric analysis. The method developed was validated in terms of its precision and accuracy. The intra- and interday precision of the marker compounds were <3.0% of the relative standard deviation (RSD) and the recovery of the marker compounds was 92.74%-104.16% with RSD values <3.0%. The results of our quantitative analysis show that the quantities of the 19 marker compounds varied between a laboratory water extract and commercial OJS granules. The chemometric analysis used, principal component analysis (PCA) and hierarchical clustering analysis (HCA), also showed that the OJS water extract produced using a laboratory method clearly differed from the commercial OJS granules; therefore, an equalized production process is required for quality control of OJS preparations. Our results suggest that the HPLC analytical methods developed are suitable for the quantification and quality assessment of OJS preparations when combined with chemometric analysis involving PCA and HCA.

Biological toxicity of groundwater in a seashore area: causal analysis and its spatial pollutant pattern

To ensure the safety of groundwater usage in a seashore area where seawater incursion and unexpected leakage are taking place, this paper utilizes the Microtox test to quantify the biological toxicity of groundwater and proposes an integrated data analysis procedure based on hierarchical cluster analysis (HCA) and principal component analysis (PCA) for determining the key environmental factors that may result in the biological toxicity, together with the spatial risk pattern associated with groundwater usage. For these reasons, this study selects the coastal area of Taichung city in Central Taiwan as an example and implements a monitoring program with 40 samples. The results indicate that the concentration of total arsenic in the coastal areas is about 0.23-270.4 μg L(-1), which is obviously higher than the interior of Taichung city. Moreover, the seawater incursion and organic pollution in the study area may be the key factors resulting in the incubation of toxic substances. The results also indicate that As(3+) is the main contributor to biological toxicity compared to other disinfection by-products. With the help of the visualized spatial pollutants pattern of groundwater, an advanced water quality control plan can be made.