Multiobjective Combinatorial Optimization Using a Single Deep Reinforcement Learning Model.

This article proposes utilizing a single deep reinforcement learning model to solve combinatorial multiobjective optimization problems. We use the well-known multiobjective traveling salesman problem (MOTSP) as an example. Our proposed method employs an encoder-decoder framework to learn the mapping from the MOTSP instance to its Pareto-optimal set. Specifically, it leverages a novel routing encoder to extract information for both the entire multiobjective aspect and every individual objective from the MOTSP instance. The global embeddings and each objective's embeddings are adaptively aggregated via a routing network to form the subproblems' embedding that can well represent the MOTSP features. Using a modified context embedding, the subproblems' embeddings are fed into a decoder to produce a set of approximate Pareto-optimal solutions in parallel. Additionally, we develop a Top-k baseline to enable more efficient data utilization and lightweight training for our proposed method. We compare our method with heuristic-based and learning-based ones on various types of MOTSP instances, and the experimental results show that our method can solve MOTSP instances in real-time and outperform the other algorithms, especially on large-scale problem instances.

Supplementation of microencapsulated probiotics modulates gut health and intestinal microbiota.

The beneficial effect of probiotics on host health is impaired due to the substantial loss of survivability during gastric transit caused by small intestinal enzymes and bile acids. Encapsulation helps to preserve the probiotics species from severe environmental factors. Lactobacillus paracasei, highly sensitive probiotic species to gastric acid, was encapsulated with polyacrylate resin. C57BL/6 male mice were equally divided into three groups; control group was fed with basal diet without any additives, the un-encapsulated group was fed with 0.1% of a mixture of encapsulating material and L. paracasei, and encapsulated group was fed with 0.1% encapsulated L. paracasei (microcapsule) for 4 weeks. The result showed elevated fecal moisture percentage in the encapsulated group, but not in the un-encapsulated group. Further study showed that the ratio of villus height to crypt depth in the small intestine was significantly higher compared to un-encapsulated and the control group. Microencapsulated probiotics also remarkably increased intestinal mucin and secretory immunoglobulin A (sIgA) concentration, intestinal MUC-2, and tight junction protein mRNA expression levels improving the intestinal barrier function of mice. In addition, microcapsules also reduced proinflammatory factor mRNA expression, while considerably increasing anti-inflammatory factor mRNA expression. Microbiota metabolites, fecal LPS (Lipopolysaccharide) were downregulated, and acetate and lactate were upraised compared to control. Furthermore, glutathione peroxidase (GSH-Px) and TAOC levels were increased and Malondialdehyde (MDA) was decreased improving antioxidant capacity. Microflora and bioinformatic predictive analysis of feces showed that encapsulated probiotics remarkably increased Lactobacillus proportions. Mice's intestinal health can thus be improved by using microencapsulated probiotics.

[Effects of long-chain noncoding RNA Linc00673 overexpression on proliferation and apoptosis of gastric cancer cell line MGC-803].

Objective: To investigate the effects of long-chain noncoding RNA Linc00673 overexpression on proliferation and apoptosis of gastric cancer cells and its mechanisms. Methods: The recombinant lentivirus expressing plasmid pLVX-Linc00673 and the control empty plasmid pLVX-NC were packaged and amplified in 293T cells, and the recombinant lentivirus was transfected into gastric cancer cell line MGC-803 to establish a cell line stably overexpressing Linc00673. The expression of Linc00673 gene was detected by real-time fluorescence quantitative PCR. The growth and proliferation of cells were observed by MTT assay and clone formation assay. Cell cycle and apoptosis were detected by flow cytometry. The expressions of cell cycle related regulatory genes were detected by qPCR. The expressions of key molecules in the PI3K/Akt signaling pathway and tumor proliferation related proteins were detected by Western blot. Results: The expressions of Linc00673 in gastric cancer cell line MGC-803, BGC-823 and AGS were significantly higher than that in normal gastric mucosa cell line GES-1 (P＜0.05). MGC-803 cell line with stable overexpression of LINC00673 was established, and the expression level of LincC00673 was 200 times higher than that of the control empty carrier group. Overexpression of Linc00673 promoted proliferation of MGC-803 cells (P＜0.05) and clone formation (P＜0.05), inhibited cell apoptosis and affected the G1→S phase progression of cell cycle (P＜0.01). Overexpression of Linc00673 could affect the expressions of cell cycle regulatory gene CCNG2, P19 and CDK1 in MGC-803. Western blot showed that Linc00673 overexpression not only promoted the expressions of the key molecule pAkt in PI3K / Akt signaling pathway and its downstream target NF-κ B and Bcl-2 protein, but also up regulated the expressions of tumor related factors ß-catenin and EZH2 proteins. Conclusion: Overexpression of Linc00673 may promote proliferation and inhibit apoptosis of MGC-803 cells through PI3K/Akt signaling pathway.

A Three-Level Radial Basis Function Method for Expensive Optimization.

This article proposes a three-level radial basis function (TLRBF)-assisted optimization algorithm for expensive optimization. It consists of three search procedures at each iteration: 1) the global exploration search is to find a solution by optimizing a global RBF approximation function subject to a distance constraint in the whole search space; 2) the subregion search is to generate a solution by minimizing an RBF approximation function in a subregion determined by fuzzy clustering; and 3) the local exploitation search is to generate a solution by solving a local RBF approximation model in the neighborhood of the current best solution. Compared with some other state-of-the-art algorithms on five commonly used scalable benchmark problems, ten CEC2015 computationally expensive problems, and a real-world airfoil design optimization problem, our proposed algorithm performs well for expensive optimization.

A Penalty-Based Differential Evolution for Multimodal Optimization.

It is very difficult to locate multiple global optimal solutions (GOSs) of multimodal optimization problems (MMOPs). To deal with this issue, a penalty-based multimodal optimization differential evolution (DE), called PMODE, is developed in this article. In PMODE, a penalty strategy with a dynamic penalty radius is constructed to solve MMOPs. An elite selection mechanism is designed to identify and select elite solutions. The neighboring areas of these elite solutions are penalized. PMODE uses a popular DE variant-JADE as its search engine. The proposed PMODE is compared with several other state-of-the-art multimodal optimization algorithms on 20 MMOPs used in the IEEE CEC2013 special session. The experimental results show that PMODE performs better than other state-of-the-art methods.

CHMP4C regulates lung squamous carcinogenesis and progression through cell cycle pathway.

BACKGROUND: Lung cancer is a common kind of human malignancies. Lung squamous cell carcinoma (LUSC) is a key subtype of lung cancer. Cell cycle plays an important role in the development and occurrence of LUSC, however, there is still a lack of cell cycle-related genes in LUSC diagnosis and prediction of prognosis. METHODS: We identified differentially expressed genes (DEGs) with "limma" package in R software, and determined the biomarkers of LUSC in diagnosing by performing receiver operating characteristic (ROC) curve analysis, the biomarker effectiveness in diagnosing LUSC was assessed by performing five-fold cross-validation with logistic regression. Kaplan-Meier plot and the nomogram assessed the relationship between the biomarker and patient survival, and WB and qRT-PCR detected the biomarker expression in cells and tissues. Flow cytometry detects the role of the biomarker in the cell cycle. RESULTS: Integration analysis with The Cancer Genome Atlas (TCGA) database obtained a unique gene related to cell cycle in LUSC (Charged multivesicular body protein 4C, CHMP4C), and the protein of CHMP4C was highly expressed in LUSC tissues. ROC analysis indicated that CHMP4C was a biomarker for the diagnosis of LUSC. Bioinformatic analysis indicated that CHMP4C might be associated with cell cycle in LUSC. CHMP4C knockdown resulted in S-phase arrest of cells with LUSC. According to the survival rate analysis, CHMP4C overexpression indicated poor prognosis in patients with LUSC. CONCLUSIONS: CHMP4C regulates the proliferation process of tumor cells through the cell cycle. It can be used as a potential diagnostic and prognostic biomarker for LUSC.

Comparison of 2 Ear Molding Systems for Nonsurgical Management of Newborn Auricular Deformities.

BACKGROUND: Congenital auricular deformities are common diseases in newborn infants. We compared the efficacy of 2 ear molding systems in the nonsurgical management of newborn auricular deformities in Chinese infants. METHODS: A total of 462 newborns with auricular deformities were treated with either the EarWell or the LiangEar ear molding systems. The posttreatment outcome was graded as excellent, fair, and poor. The differences in effectiveness between the 2 ear molding systems and factors that may affect the posttreatment outcome were analyzed. RESULTS: Both ear molding systems showed substantial efficacy in the treatment of newborn auricular deformities. The effective rate was comparable between the EarWell and the LiangEar systems for 4 types of auricular deformities (cryptotia, prominent ear, helical rim abnormality, and cup ear), while the costs for the LiangEar systems were half as much as that for the EarWell systems. Multivariate regression analysis showed that earlier time points at treatment initiation, less severe auricular deformities, treatment duration, and breastfeeding were the most significant predictive factors for a better outcome. CONCLUSION: Our findings demonstrate that using the EarWell and the LiangEar systems are both optimal nonsurgical approaches for treatment of most newborn auricular deformities.