Integration of protein L-immobilized epoxy magnetic bead capture with LC-MS/MS for therapeutic monoclonal antibody quantification in serum.

In recent years, there has been a growing interest in the thriving monoclonal antibody (mAb) industry due to the wide utilization of mAbs in clinical therapies. Robust and accurate bioanalytical methods are required to enable fast quantification of mAbs in biological matrices, especially in the context of pharmacokinetics (PKs)/pharmacodynamics (PDs) and therapeutic drug monitoring (TDM) studies. In this investigation, we presented a novel immuno-magnetic capture coupled with a liquid chromatography-tandem mass spectrometry (LC-MS/MS) method designed for the quantification of immunoglobulin G-kappa-based mAbs in biological fluids. The immunoaffinity absorbent for mAb drug purification was meticulously crafted by immobilizing protein L onto monosize, magnetic poly(glycidyl methacrylate) (m-pGMA) beads, synthesized through dispersion polymerization. The microspheres were acquired with an average size of 1.6 µm, and the optimal binding of mAbs from the aqueous mAb solution was determined to be 45.82 mg g-1. The quantification of mAbs in 10 µL serum samples was achieved through affinity purification using m-pGMA@protein L beads (employing rituximab as an internal standard (IS)), on-bead reduction, and rapid tryptic digestion. Remarkably, the entire process, taking less than 2.5 hours, held significant potential for simplifying pretreatment procedures and minimizing analytical time. Furthermore, the developed method underwent validation in accordance with the European Medicines Agency (EMA) guidelines. The assay demonstrated commendable linearity within the 2-400 µg mL-1 range for both daratumumab and pembrolizumab. Intra- and inter-assay coefficients of variation fell within the range of 0.7% to 13.4%, meeting established acceptance criteria. Other validation parameters also conformed to regulatory standards. Ultimately, the efficacy of the method was substantiated in a pharmacokinetic study following a single-dose intravenous administration to mice, underscoring its applicability and reliability in real-world scenarios.

Deep Probabilistic Principal Component Analysis for Process Monitoring.

Probabilistic latent variable models (PLVMs), such as probabilistic principal component analysis (PPCA), are widely employed in process monitoring and fault detection of industrial processes. This article proposes a novel deep PPCA (DePPCA) model, which has the advantages of both probabilistic modeling and deep learning. The construction of DePPCA includes a greedy layer-wise pretraining phase and a unified end-to-end fine-tuning phase. The former establishes a hierarchical deep structure based on cascading multiple layers of the PPCA module to extract high-level features. The latter builds an end-to-end connection between the raw inputs and the final outputs to further improve the representation of the model to high-level features. After constructing the model structure of DePPCA, we first present the detailed training processes of the pretraining and fine-tuning stages, then clarify the theoretical merits of the proposed model from the perspective of variational inference. For process monitoring purposes, we develop two statistics based on the established DePPCA. The monitoring performance of these two statistics can remain superior even if the features extracted by DePPCA are significantly compressed to univariate. This makes the feature extraction process and online monitoring procedure of DePPCA quite fast. In other words, the proposed DePPCA can achieve accurate and efficient process monitoring by only extracting one feature for each sample. Finally, the effectiveness of DePPCA is evaluated on the Tennessee Eastman (TE) process and the multiphase flow (MPF) facility.

BRISC is required for optimal activation of NF-κB in Kupffer cells induced by LPS and contributes to acute liver injury.

BRISC (BRCC3 isopeptidase complex) is a deubiquitinating enzyme that has been linked with inflammatory processes, but its role in liver diseases and the underlying mechanism are unknown. Here, we investigated the pathophysiological role of BRISC in acute liver failure using a mice model induced by D-galactosamine (D-GalN) plus lipopolysaccharide (LPS). We found that the expression of BRISC components was dramatically increased in kupffer cells (KCs) upon LPS treatment in vitro or by the injection of LPS in D-GalN-sensitized mice. D-GalN plus LPS-induced liver damage and mortality in global BRISC-null mice were markedly attenuated, which was accompanied by impaired hepatocyte death and hepatic inflammation response. Constantly, treatment with thiolutin, a potent BRISC inhibitor, remarkably alleviated D-GalN/LPS-induced liver injury in mice. By using bone marrow-reconstituted chimeric mice and cell-specific BRISC-deficient mice, we demonstrated that KCs are the key effector cells responsible for protection against D-GalN/LPS-induced liver injury in BRISC-deficient mice. Mechanistically, we found that hepatic and circulating levels of TNF-α, IL-6, MCP-1, and IL-1ß, as well as TNF-α- and MCP-1-producing KCs, in BRISC-deleted mice were dramatically decreased as early as 1 h after D-GalN/LPS challenge, which occurred prior to the elevation of the liver injury markers. Moreover, LPS-induced proinflammatory cytokines production in KCs was significantly diminished by BRISC deficiency in vitro, which was accompanied by potently attenuated NF-κB activation. Restoration of NF-κB activation by two small molecular activators of NF-κB p65 effectively reversed the suppression of cytokines production in ABRO1-deficient KCs by LPS. In conclusion, BRISC is required for optimal activation of NF-κB-mediated proinflammatory cytokines production in LPS-treated KCs and contributes to acute liver injury. This study opens the possibility to develop new strategies for the inhibition of KCs-driven inflammation in liver diseases.

Insights into the mechanisms involved in the fungal degradation of plastics.

Fungi are considered among the most efficient microbial degraders of plastics, as they produce salient enzymes and can survive on recalcitrant compounds with limited nutrients. In recent years, studies have reported numerous species of fungi that can degrade different types of plastics, yet there remain many gaps in our understanding of the processes involved in biodegradation. In addition, many unknowns need to be resolved regarding the fungal enzymes responsible for plastic fragmentation and the regulatory mechanisms which fungi use to hydrolyse, assimilate and mineralize synthetic plastics. This review aims to detail the main methods used in plastic hydrolysis by fungi, key enzymatic and molecular mechanisms, chemical agents that enhance the enzymatic breakdown of plastics, and viable industrial applications. Considering that polymers such as lignin, bioplastics, phenolics, and other petroleum-based compounds exhibit closely related characteristics in terms of hydrophobicity and structure, and are degraded by similar fungal enzymes as plastics, we have reasoned that genes that have been reported to regulate the biodegradation of these compounds or their homologs could equally be involved in the regulation of plastic degrading enzymes in fungi. Thus, this review highlights and provides insight into some of the most likely regulatory mechanisms by which fungi degrade plastics, target enzymes, genes, and transcription factors involved in the process, as well as key limitations to industrial upscaling of plastic biodegradation and biological approaches that can be employed to overcome these challenges.

Security Versus Accuracy: Trade-Off Data Modeling to Safe Fault Classification Systems.

While the data-driven fault classification systems have achieved great success and been widely deployed, machine-learning-based models have recently been shown to be unsafe and vulnerable to tiny perturbations, i.e., adversarial attack. For the safety-critical industrial scenarios, the adversarial security (i.e., adversarial robustness) of the fault system should be taken into serious consideration. However, security and accuracy are intrinsically conflicting, which is a trade-off issue. In this article, we first study this new trade-off issue in the design of fault classification models and solve it from a brand new view, hyperparameter optimization (HPO). Meanwhile, to reduce the computational expense of HPO, we propose a new multiobjective (MO), multifidelity (MF) Bayesian optimization (BO) algorithm, MMTPE. The proposed algorithm is evaluated on safety-critical industrial datasets with the mainstream machine learning (ML) models. The results show that the following hold: 1) MMTPE is superior to other advanced optimization algorithms in both efficiency and performance and 2) fault classification models with optimized hyperparameters are competitive with advanced adversarially defensive methods. Moreover, insights into the model security are given, including the model intrinsic security properties and the correlations between hyperparameters and security.

Deep PLS: A Lightweight Deep Learning Model for Interpretable and Efficient Data Analytics.

The salient progress of deep learning is accompanied by nonnegligible deficiencies, such as: 1) interpretability problem; 2) requirement for large data amounts; 3) hard to design and tune parameters; and 4) heavy computation complexity. Despite the remarkable achievements of neural networks-based deep models in many fields, the practical applications of deep learning are still limited by these shortcomings. This article proposes a new concept called the lightweight deep model (LDM). LDM absorbs the useful ideas of deep learning and overcomes their shortcomings to a certain extent. We explore the idea of LDM from the perspective of partial least squares (PLS) by constructing a deep PLS (DPLS) model. The feasibility and merits of DPLS are proved theoretically, after that, DPLS is further generalized to a more common form (GDPLS) by adding a nonlinear mapping layer between two cascaded PLS layers in the model structure. The superiority of DPLS and GDPLS is demonstrated through four practical cases involving two regression problems and two classification tasks, in which our model not only achieves competitive performance compared with existing neural networks-based deep models but also is proven to be a more interpretable and efficient method, and we know exactly how it improves performance, how it gives correct results. Note that our proposed model can only be regarded as an alternative to fully connected neural networks at present and cannot completely replace the mature deep vision or language models.

Predictive Modeling With Multiresolution Pyramid VAE and Industrial Soft Sensor Applications.

In industrial processes, the sampling rates of process variables are discrepant because of the nature of instruments and measuring demands, which forms the challenging issue, that is, the multirate modeling in the data-driven soft sensor development. In this work, a multiresolution pyramid variational autoencoder (MR-PVAE) predictive model is proposed to solve this problem based on the deep feature extraction and feature pyramid augmentation. First, a multirate data filter is designed through a resolution searching strategy to turn the original process data into a multiresolution dataset. Then, the pyramid variational autoencoder (PVAE) is proposed to extract deep nonlinear features from the data with different resolutions. In PVAE, the augmented feature pyramid is constructed layer by layer to fuse extracted features from low resolution to the high. As a consequence, the extracted features with various resolutions are gathered to form the regression model, where the process information contained in data with discrepant sampling rates can be fully utilized. Due to the layer-by-layer enhanced features, the prediction accuracy of the soft sensing model are gradually improved. Meanwhile, an optimized training strategy is established to select the optimal feature pyramid for prediction. A numerical experiment and an industrial soft sensing case are given to validate the effectiveness and superiority of the proposed MR-PVAE model.

Long-term hypoxia inhibits the passage-dependent stemness decrease and senescence increase of human dental pulp stem cells.

Dental pulp stem cells (DPSCs) derived from discarded orthodontic teeth are easily obtained and have become a promising source for mesenchymal stem cell-based therapy. However, the pulp tissue is limited, and long-term culture induces cell senescence. Hypoxic culture was expected to be suitable for DPSC expansion, but the results have been contradictory. The aim of this study was to verify the effect of hypoxic culture on human DPSCs (hDPSCs). hDPSCs were isolated and cultured in normoxic (ambient O2 concentration) and hypoxic (5% O2) environments from passage 3 (P3) to P6. The biological characteristics of the cells at P4 (short-term culture) and P6 (long-term culture) were evaluated, including the expression of surface markers, cellular proliferation activity, cellular senescence, and spontaneous and induced differentiation. The results showed that the morphology, phenotype, and proliferation activity of hDPSCs were not affected by hypoxic culture. Long-term normoxic culture of hDPSCs induced cell stemness loss and cell senescence, while hypoxic culture could alleviate these effects. The expression of the stemness markers STRO-1 and OCT4 was increased and the number of senescent cells and the expression of the senescence-related genes P53 and TGF-ß were reduced by long-term hypoxic culture. Spontaneous osteogenic and adipogenic differentiation did not occur during long-term normoxic culture. However, hypoxic culture suppressed the expression of the osteogenic markers ALP and RUNX-2 and the adipogenic markers PPAR-γ and FABP4. The induced osteogenic and adipogenic differentiation was apparently reduced by hypoxic culture as well. Our findings indicate that long-term hypoxia culture is beneficial to the maintenance of hDPSCs' biological characteristics and provide some insights into their large-scale expansion.

Gated Stacked Target-Related Autoencoder: A Novel Deep Feature Extraction and Layerwise Ensemble Method for Industrial Soft Sensor Application.

These days, data-driven soft sensors have been widely applied to estimate the difficult-to-measure quality variables in the industrial process. How to extract effective feature representations from complex process data is still the difficult and hot spot in the soft sensing application field. Deep learning (DL), which has made great progresses in many fields recently, has been used for process monitoring and quality prediction purposes for its outstanding nonlinear modeling and feature extraction abilities. In this work, deep stacked autoencoder (SAE) is introduced to construct a soft sensor model. Nevertheless, conventional SAE-based methods do not take information related to target values in the pretraining stage and just use the feature representations in the last hidden layer for final prediction. To this end, a novel gated stacked target-related autoencoder (GSTAE) is proposed for improving modeling performance in view of the above two issues. By adding prediction errors of target values into the loss function when executing a layerwise pretraining procedure, the target-related information is used to guide the feature learning process. Besides, gated neurons are utilized to control the information flow from different layers to the final output neuron that take full advantage of different levels of abstraction representations and quantify their contributions. Finally, the effectiveness and feasibility of the proposed approach are verified in two real industrial cases.

Overexpression of Hepatocyte Growth Factor in Dental Pulp Stem Cells Ameliorates the Severity of Psoriasis by Reducing Inflammatory Responses.

Psoriasis is an autoimmune disease still lacking standard treatment, and it has been demonstrated that mesenchymal stem cells (MSCs) are capable of immunoregulation. The underlying mechanism might involve the secretion of soluble cytokines, such as hepatocyte growth factor (HGF). This study aims to investigate the therapeutic effect of HGF-overexpressed dental pulp stem cells (DPSCs) [DPSCs; HGF overexpressed DPSCs (HGF-DPSCs)] on imiquimod-induced psoriasis. DPSCs were isolated and transfected by adenovirus vector carrying HGF gene (Ad-HGF). The immunoregulatry abilities of DPSCs and HGF-DPSCs were investigated by coculture of the MSCs with peripheral blood mononuclear cells (PBMCs) under appropriated stimulation. The psoriatic mice were treated with saline control, DPSCs, or HGF-DPSCs. Then the mice spleens were collected and weighted. The psoriatic skin lesions were analyzed by Hematoxylin/Eosin and immunohistochemical staining for histopathological changes, and quantitative real-time polymerase chain reaction to detect the expression levels of CD4+ T cell-related transcription factors and cytokines. The mice blood serum was measured by MILLIPLEX analysis and enzyme-linked immunosorbent assay to evaluate the expression levels of inflammation cytokines. The coculture experiments showed HGF overexpression enhanced the immunoregulation abilities of DPSCs not by suppressing PBMCs' proliferation, but by downregulating T helper 1 (Th1), Th17 cells, and upregulating regulatory T (Treg) cells. In psoriatic skin lesions, the psoriasis-like erythema, scaling, and thickening were ameliorated; and the expression of cytokeratin 6 (CK6), and cytokeratin 17 (CK17) were downregulated by DPSCs and HGF-DPSCs treatment. HGF overexpression enhanced the decrease of spleen masses; enhanced the downregulation of the expression levels of interferon-gamma (IFN-γ), tumor necrosis factor-α, and interleukin (IL)-17A in the blood serums; enhanced the downregulation of T-box transcription factor 21 (T-bet), IFN-γ, retinoic acid-related orphan receptor-γt (RORγt), IL-17A, IL-17F, IL-23, and upregulation of Foxp3 and IL-10 in the psoriatic skin lesions. Therefore, HGF overexpression enhanced DPSCs' treatment effect on psoriasis mainly by reducing inflammatory responses. These findings might provide new immunoregulation strategies for psoriasis treatment.

Hierarchical Quality Monitoring for Large-Scale Industrial Plants With Big Process Data.

For large-scale industrial plants, quality-related process monitoring is challenging because of the complex features of multiunit, multimode, high-dimension data. Hence, a hierarchical quality monitoring (HQM) algorithm based on the distributed parallel semisupervised Gaussian mixture model (dp-S2GMM) is proposed in this article. In HQM, a large-scale process is first decomposed into a group of unit blocks according to the process structure. Subsequently, in each block, a quality regression model with multimode big process data is built using the dp-S2GMM, which is derived from a scalable stochastic variational inference semisupervised GMM (SVI-S2GMM). With the regression model, a hierarchical fault detection and diagnosis scheme in both quality-related and quality-unrelated subspaces is proposed from the variable level, block level to plant-wide level. Finally, an industrial case study on the Tennessee Eastman process demonstrates the feasibility and effectiveness of the proposed HQM algorithm.

Semisupervised Bayesian Gaussian Mixture Models for Non-Gaussian Soft Sensor.

Soft sensors have been widely accepted for online estimating key quality-related variables in industrial processes. The Gaussian mixture models (GMM) is one of the most popular soft sensing methods for the non-Gaussian industrial processes. However, in industrial applications, the quantity of samples with known labels is usually quite limited because of the technical limitations or economical reasons. Traditional GMM-based soft sensor models solely depending on labeled samples may easily suffer from singular covariances, overfitting, and difficulties in model selection, which results in the performance deterioration. To tackle these issues, we propose a semisupervised Bayesian GMM (S2BGMM). In the S2BGMM, we first propose a semisupervised fully Bayesian model, which enables learning from both the labeled and unlabeled datasets for remedying the deficiency of infrequent labeled samples. Subsequently, a general framework of weighted variational inference is developed to train the S2BGMM, such that the rate of learning from unlabeled samples can be controlled by penalizing the unlabeled dataset. Case studies are carried out to evaluate the performance of the S2BGMM through a numerical example and two real-world industrial processes, which demonstrate the effectiveness and reliability of the proposed approach.

Simultaneous quantitative determination of liraglutide and insulin degludec in rat plasma by liquid chromatography-tandem mass spectrometry method and its application.

A simple, fast and high-throughput LC-tandem mass spectrometry method was developed and validated to simultaneously measure liraglutide and insulin degludec in rat plasma. After protein precipitation, plasma samples were subjected to gradient elution using an InertSustain Bio C18 column with 1000/20/1 water/acetonitrile/formic acid (v/v/v) and 1000/1 acetonitrile/formic acid (v/v) as the mobile phase. The method was validated from 1.00 to 500 ng/mL of liraglutide and insulin degludec. Further, the extraction recovery from the plasma was 41.8%-49.2% for liraglutide and 56.5%-69.7% for insulin degludec. Intra- and inter-day precision of liraglutide was 3.5%-9.4% and 8.4%-9.8%, respectively, whereas its accuracy was between -12.6% and -1.3%. Intra- and inter-day precision of insulin degludec was 5.2%-13.6% and 11.8%-19.1%, respectively, showing an accuracy between -3.0% and 9.9%. As a result, the method was successfully applied to a pharmacokinetics study of liraglutide and insulin degludec following a single-dose subcutaneous administration to rats.

Method development and validation of LC-MS/MS-based assay for the simultaneous quantitation of trastuzumab and pertuzumab in cynomolgus monkey serum and its application in pharmacokinetic study.

We present a simple and robust LC-MS/MS assay for the simultaneous quantitation of an antibody cocktail of trastuzumab and pertuzumab in monkey serum. The LC-MS/MS method saved costs, decreased the analysis time, and reduced quantitative times relative to the traditional ligand-binding assays. The serum samples were digested with trypsin at 50°C for 60 min after methanol precipitation, ammonium bicarbonate denaturation, dithiothreitol reduction, and iodoacetamide alkylation. The tryptic peptides were chromatographically separated using a C18 column (2.1 × 50 mm, 2.6 µm) with mobile phases of 0.1% formic acid in water and acetonitrile. The other monoclonal antibody, infliximab, was used as internal standards to minimize the variability during sample processing and detection. A unique peptide for each monoclonal antibody was simultaneously quantified using LC-MS/MS in the multiple reaction monitoring mode. Calibration curves were linear from 2.0 to 400 µg/mL. The intra- and inter-assay precision (%CV) was within 8.9 and 7.4% (except 10.4 and 15.1% for lower limit of quantitation), respectively, and the accuracy (%Dev) was within ±13.1%. The other validation parameters were evaluated, and all results met the acceptance criteria of the international guiding principles. Finally, the method was successfully applied to a pharmacokinetics study after a single-dose intravenous drip administration to cynomolgus monkeys.

Magnetic nanoparticles encapsulated laccase nanoflowers: evaluation of enzymatic activity and reusability for degradation of malachite green.

Magnetic laccase nanoflowers (MNFs-Lac) were successfully prepared through encapsulating Fe3O4 magnetic nanoparticles into the interior of laccase nanoflowers by grafting N-(phosphonomethyl)iminodiacetic acid (PMIDA) as an interconnecting bridge between the magnetic nanoparticles and copper ions. The characterizations by scanning electron microscopy and transmission electron microscopy showed that MNFs-Lac were spherical, porous and flower-like crystals with diameters of ∼10 µm, and Fe3O4 nanoparticles were encapsulated in the interior of MNFs-Lac evenly. The enzymatic activity and reusability of MNFs-Lac were evaluated based on the degradation efficiency for malachite green (MG). The degradation parameters, concerning initial MG concentration, dosage of MNFs-Lac, reaction temperature, pH value and reaction time, were optimized through single-factor experiments. Under the optimal conditions, 25 mg·L-1 MG can be degraded almost completely by 1.5 g·L-1 MNFs-Lac within 15 min. When the MNFs-Lac were reused for 18 times, the degradation efficiency of MG was still as high as 90%. These results suggested that the modified preparation method improved greatly the reusability of MNFs-Lac, which made them more suitable to degrade MG in a water environment.

A Sandwich Structure of Human Dental Pulp Stem Cell Sheet, Treated Dentin Matrix, and Matrigel for Tooth Root Regeneration.

Tooth loss can cause a lot of physiological and psychological suffering. And tooth root engineering is a promising way for tooth loss treatment. Two kinds of seed cells are usually adopted for tooth root regeneration. In this study, a practical sandwich structure for tooth root regeneration was developed, which was constituted by only one kind of seed cell: human dental pulp stem cells (hDPSCs) and three kinds of graft materials: Vitamin C (VC) induced hDPSC sheet, human treated dentin matrix (hTDM), and Matrigel. It was found that VC could induce hDPSCs to form a cell sheet with two or three cell layers and promote their collagen type I (COL1) mRNA expression obviously. hDPSCs could attach and grow on hTDM, and the mRNA expression of osteocalcin (OCN), dentin sialophosphoprotein (DSPP), vascular endothelial growth factor receptor 1 (VEGFR1), and Nestin in hDPSCs was obviously upregulated by hTDM leaching solution. hDPSCs could stretch and proliferate in Matrigel. And when cultured in Matrigel condition medium, they positively expressed CD31, ß3-Tubulin, and Nestin proteins, as well as increased the mRNA expression of OCN, ALP, and Nestin. Furthermore, periodontium, dentin, and pulp-like tissues were successfully regenerated after the sandwich structure of hDPSC sheet/TDM/Matrigel was transplanted in nude mice subcutaneously for 3 months. Periodontium-like dense connective tissue was regenerated around the hTDM, and a great mass of predentin was formed on the cavity side of hTDM. Odontoblast-like cells and blood vessel-like structures, even nerve-like fibers, were observed in the pulp cavity. In summary, the above results showed that hDPSCs could be used as seed cells for the whole tooth root regeneration, and the sandwich structure constituted by hDPSC sheet, TDM/hDPSCs, and Matrigel/hDPSCs could be utilized for tooth root regeneration.

Removal of doxorubicin by magnetic copper phosphate nanoflowers for individual urine source separation.

Doxorubicin (DOX) originated from users' urine has been an emerging environmental pollutant due to its significant genotoxicity to mankind. Thus, urine source separation is a potential strategy to isolate DOX at a higher concentration and reduce the burden of downstream wastewater treatment. To develop highly efficient, easy separation and retrievable materials for individual patient to conveniently remove DOX from own urine, magnetic Cu3(PO4)2 nanoflowers were prepared through anchoring amino-functionalized magnetic nanoparticles on the Cu3(PO4)2 nanoflowers. Characterizations revealed the magnetic nanoflowers were spherical in shape with a mean size of 15 µm, and porous and hierarchical in structure. Magnetic nanoparticles located the surface of petals. Multibatch experiments were performed to assess the removal performance of DOX from aqueous solution. The magnetic nanoflowers exhibited excellent removal efficiency of DOX under weakly alkaline condition at ambient temperature. Linear and non-linear analyses were carried out to compare the best fitting kinetics and isotherms. Sorption kinetic data best fitted the pseudo-second order model. The Freundlich isotherm explained equilibrium sorption data with R2 = 0.993 higher than that for the Langmuir isotherm. When the pH of synthetic urine was adjusted to weakly alkaline (pH 8.0-9.0), over 95% of DOX (20 mg L-1) was removed by a little of magnetic nanoflowers (50 mg L-1) within 5 min. Meanwhile, the magnetic nanoflowers could be easily separated and recovered from the synthetic urine by a magnet. So, for individual urine source separation strategy, the magnetic nanoflower seems to be an efficient, convenient and inexpensive approach to remove DOX from human urine.

Ensemble semi-supervised Fisher discriminant analysis model for fault classification in industrial processes.

In this paper, an ensemble form of the semi-supervised Fisher Discriminant Analysis (FDA) model is developed for fault classification in industrial processes. This method uses the K Nearest Neighbor (KNN) algorithm to merge the metric level outputs, which are obtained by the sub-classifiers in the ensemble model, to get the final classification result. An adaptive form is further proposed to improve the classification performance by putting forward to a new method of weight adjustment. While semi-supervised learning can generate a better model by exploiting additional information contained in unlabeled data, ensemble learning achieves the promotion of algorithm robustness by integrating a series of weak learners. In addition, the property of diversity in ensemble learning can be boosted by incorporating different unlabeled data to different weak learners. Therefore, the combination of those two methods can achieve great generalization for the fault classification model. The performances of two proposed methods are evaluated through an industrial benchmark process.

Evaluation of Nile Red-Loaded Mesoporous Silica Nanoparticles for Developing Water-Soaked Fingerprints on Thermal Paper.

Nile red has been an alternative reagent for detecting latent fingerprints on wetted substrates. However, the presence of methanol in nile red solution could make injury to handlers and destroy the traces on surfaces, such as texts on thermal papers. A novel small particle reagent formulation constituting of mesoporous silica nanoparticles (MSNs) based on nile red was prepared to overcome the problem. Compared with the conventional reagents Oil Red O or nile red solution, the nile red-loaded MSNs are highly selective to lipid residues of fingerprints and showed a greater ability to develop clear, sharp, and detailed fingerprints on thermal papers after these were immersed in water. In addition, it can retain texts on the thermal papers well and use only water as a solvent. These suggested that nile red-loaded MSNs are a safe, efficient, and convenient method to develop latent fingerprints on wide range of substrates of forensic importance.

Efficient removal of ethidium bromide from aqueous solution by using DNA-loaded Fe3O4 nanoparticles.

Ethidium bromide (EtBr) is widely used as DNA-staining dyes for the detection of nucleic acids in laboratories and known to be powerful mutagens and carcinogens. In the present paper, the removal of EtBr from aqueous solutions in a batch system using DNA-loaded Fe3O4 nanoparticles as a simple and efficient method was investigated. DNA was covalently loaded on the surface of Fe3O4 magnetic nanoparticles, which was confirmed by FT-IR analysis and zeta potential measurements. The morphology and crystal structure were characterized by SEM, TEM, and XRD. The influence factors on the removal efficiency such as initial EtBr concentration, contact time, adsorbent dose, pH, and temperature were also studied. The removal process of EtBr can be completed quickly within 1 min. The removal efficiency was more than 99% while the EtBr concentration was routinely used (0.5 mg L-1) in biology laboratories and the dosages of nanoparticles were 1 g L-1. For the different EtBr concentrations from 0.5 to 10 mg L-1 in aqueous solution, the goal of optimized removal was achieved by adjusting the dosage of DNA-loaded Fe3O4 nanoparticles. The optimum pH was around 7 and the operational temperature from 4 to 35 °C was appropriate. Kinetic studies confirmed that the adsorption followed second-order reaction kinetics. Thermodynamic data revealed that the process was spontaneous and exothermic. The adsorption of EtBr on DNA-loaded Fe3O4 nanoparticles fitted well with the Freundlich isotherm model. These results indicated that DNA-loaded Fe3O4 nanoparticles are a promising adsorbent for highly efficient removal of EtBr from aqueous solution in practice.