Journal of Chemometrics最新文献

Immersive analytics is a developing field growing as technology improves. This paper presents some important points, but by no means is the discussion complete. The cited papers and books should be read to fully grasp the potential of the general field of immersive analytics. The direction of this paper is to highlight those components useful for chemometric data analyses in virtual reality.

This study demonstrates that integrating infrared spectroscopy with machine learning enables highly accurate, nondestructive classification of red-stamp ink manufacturers. We evaluated five classifiers—partial least squares discriminant analysis (PLS-DA), k-nearest neighbor (k-NN), support vector machine (SVM), random forest (RF), and a feed-forward neural network (FNN)—across multiple spectral regions. The FNN trained on second-derivative spectra in the 1700–900 cm⁻¹ region achieved perfect test metrics (F1 = 1.000; AUC = 1.000), while PLS-DA and RF also performed robustly (F1 ≥ 0.933). Variable importance in projection (VIP) analysis identified the 1650–1100 cm⁻¹ subrange as most informative, streamlining feature selection and model training. Applied to three unknown samples, the optimized FNN produced high-confidence manufacturer predictions consistent with expected origins. These results confirm that targeted spectral selection combined with derivative preprocessing markedly enhances nondestructive ink classification for forensic applications.

The present study aims to find azole-containing acetylcholinesterase (AChE) inhibitors for the treatment of Alzheimer's disease (AD) through a mixed in silico approach. The first step involved the collection of azole derivatives and predictive quantitative structure–activity relationship (QSAR) model development for their AChE inhibition activity, using multiple linear regressions (MLRs) with the genetic algorithm (GA) for feature selection. The developed GA-MLR models were statistically robust enough internally (R²_adj = 0.643–0.640, Q²_LOO = 0.616–0.621) as well as externally (R²_pred = 0.626–0.658, R²M = 0.562–0.601). The prediction reliability of the models was assured through the leverage approach of the applicability domain. The most significant models were applied to azole-containing PubChem database compounds, which were classified as active and inactive based on theoretical predictions. The toxicity analysis was also performed for the active compounds by the online web server Protox-II. The less or nontoxic compounds were subjected to molecular docking, along with donepezil as a standard. Docking analysis revealed that the four compounds have better binding affinity (binding energy = −11.6 to −11.2 kcal/mol) as compared to donepezil (binding energy = −11 kcal/mol). Apart from binding energy, donepezil was observed to be toxic by the prediction from the Protox-II. Finally, molecular dynamics (MD) analysis of two compounds (Compound 5, having the lowest IC₅₀, and Compound 25, having the highest IC₅₀ among the top 4 docked compounds) not only differentiated them based on final interactions but also exhibited that the toxicity of donepezil might be due to hydrogen bonding with the active site.

A method is introduced to perform simultaneous sparse dimension reduction on two blocks of variables. Beyond dimension reduction, it also yields an estimator for multivariate regression with the capability to intrinsically deselect uninformative variables in both independent and dependent blocks. An algorithm is provided that leads to a straightforward implementation of the method. The benefits of simultaneous sparse dimension reduction are shown to carry through to enhanced capability to predict a set of multivariate dependent variables jointly. Both in a simulation study and in two chemometric applications, the new method outperforms its dense counterpart, as well as multivariate partial least squares.

Multivariate extensions of repeated measures linear mixed models, such as repeated measures ANOVA simultaneous component analysis (RM-ASCA+) and linear mixed model-principal component analysis (LiMM-PCA), can be used for analyzing longitudinal studies with multivariate outcomes. However, there are no gold standards to assess the statistical validation of the observed effects of such models. Using real and simulated data, we here perform an empirical comparison of different strategies for assessing statistical significance in these frameworks: permutation tests, the global log-likelihood ratio (GLLR) test, and nonparametric bootstrap confidence intervals for the estimated multivariate effects. Power curves were used to examine the statistical power of the different tests in detecting time–treatment interactions with varying effect sizes. Our results show that both the permutation tests and the GLLR test can be used to statistically test the presence of a time–treatment interaction effect for multivariate data; however, the GLLR approach will be sensitive to the number of included principal components in LiMM-PCA. The bootstrap confidence interval approach generally shows good statistical power but has inflated Type 1 error rates under certain conditions. This makes it unsuitable for the purpose of hypothesis testing in its present implementation, although it may still be useful for exploratory purposes. Overall, our results show that the power of the tests for assessing multivariate effects in longitudinal studies is dependent on characteristics of the dataset, and it is important to be aware of the strengths and weaknesses of the different validation procedures.

An intelligent and accurate modeling method is proposed combining near-infrared (NIR) spectroscopy for measuring organic matter content in soil samples. The proposed method uses Monte Carlo (MC) random sampling in the training set, where subsets were randomly selected from the samples and further selected using the butterfly optimization algorithm (BOA) to construct partial least squares (PLS) submodels, named MC-BOA-PLS. Ultimately, the final prediction was obtained by averaging the predictions of these submodels. The parameters of the MC-BOA-PLS model were optimized, including the iteration number of BOA, the number of butterflies, and the number of PLS submodels. Results show that MC-BOA-PLS exhibited superior predictive performance to predict organic matter content in soil compared with PLS and BOA-PLS.