Search

Additional file 1: Figure S1. The protein levels and Hazard Ratio of MARCH1-11 in TCGA-COAD, TCGA-READ, GSE12945 and GSE39582 datasets. A The protein levels of MARCH1-11 in TCGA-COAD, TCGA-READ, and GSE39582 datasets. B The forestplots of MARCH1-11 in TCGA-READ, GSE12945 and GSE39582 datasets. Figure S2. The expression levels and overall survival of MARCH8 in public datasets. A The expression levels of MARCH8 in TCGA-COAD, TCGA-READ, GSE41258 and GSE39582 datasets. B The overall survival of MARCH8 in TCGA-COAD, GSE41258, GSE12945, GSE39582, GSE17537, GSE14333 and GSE17536 datasets. Figure. S3. Reduced expression of MARCH8 in CRC tumor samples is associated with poor prognosis in patients with colorectal cancer (CRC). A Representative images of immunohistochemical analysis of the tumor samples obtained from the 85 patients with CRC. Based on the IHC analysis, the patients were divided into MARCH8-high and MARCH8-low groups. Scale bar: 100 μm. B MARCH8 mRNA levels in 25 CRC tumor and 25 paired normal tissue samples, which were obtained from the 85 patients with CRC. C Five-year survival rates of the 34 patients in the MARCH8-high group and the 51 patients in the MARCH8 low group. Figure S4. MARCH8 protein levels are negatively correlated with that of HK2 in human CRC tumor tissues. A Protein levels of MARCH8 and HK2 in 16 CRC tumor tissues and 8 normal colon tissues. B Densitometrial analysis on the ratios of MARCH8 and HK2 band. C Pearson correlation analysis of the correlation between HK2 protein levels and MARCH8. Table S1. Demographic and pathological data of the 85 patients with CRC, who were included in this study. Table S2. Short hairpin sequences of MARCH8 in this study. Table S3. Primer sequences used in this study. Table S4. Correlation of MARCH8 expression in colorectal cancer tissues with different clinicopathological features (n = 85).

Adversarial attacking aims to fool deep neural networks with adversarial examples. In the field of natural language processing, various textual adversarial attack models have been proposed, varying in the accessibility to the victim model. Among them, the attack models that only require the output of the victim model are more fit for real-world situations of adversarial attacking. However, to achieve high attack performance, these models usually need to query the victim model too many times, which is neither efficient nor viable in practice. To tackle this problem, we propose a reinforcement learning based attack model, which can learn from attack history and launch attacks more efficiently. In experiments, we evaluate our model by attacking several state-of-the-art models on the benchmark datasets of multiple tasks including sentiment analysis, text classification and natural language inference. Experimental results demonstrate that our model consistently achieves both better attack performance and higher efficiency than recently proposed baseline methods. We also find our attack model can bring more robustness improvement to the victim model by adversarial training. All the code and data of this paper will be made public., Comment: work in progress, 10 pages, 6 figures

This paper presents an improved genetic algorithm to solve the agile earth observing satellite mission planning problem. We study how to rapidly generate high quality initial solutions, and four generation strategies are proposed. The effect of the settings of operator parameters on the performance of the algorithm is analyzed. The experiment results show that the genetic algorithm based on high quality initial solutions generated by Hybrid Random Heuristic Strategy (HRHS) is more effective in solving the agile satellite mission planning problem, but in a certain time cost. We expect that our results will provide insights for the future application of genetic algorithm to satellites mission planning problems.

Augmentation of pedicle screws with bioactive glass (BG) was performed in osteoporotic ovine spine in vivo. Biomechanical tests, micro-computed tomography (CT) analysis, and histologic observation were performed.To investigate the biomechanical stability of pedicle screws augmented by BG in osteoporotic sheep and observe the bone-screw interface histologically.There is little information on the long-term biomechanical performance and screw-bone interfacial bonding of pedicle screws augmented with BG in osteoporotic spine in vivo.Twelve months after ovariectomy combined with methylprednisolone injection, 8 adult female sheep were randomly divided into 2 groups (3- and 6-mo time point groups). In each time point group, pedicles were randomly selected from the lumbar spine (L1-L6) and implanted with (1) pedicle screw alone; (2) pedicle screw augmented by polymethylmethacrylate; or (3) pedicle screw augmented by BG. Three and 6 months after implantation, animals were labeled with tetracycline and calcein before being killed. Then vertebrae with pedicle screws were obtained, and a micro-CT scan, histologic analysis, and biomechanical tests were performed.Three months after implantation, micro-CT reconstruction showed that microstructural parameters of the BG group were significantly better compared with those in the other 2 groups (P0.05). Histologic observation revealed that bone trabeculae around the screws in the BG group were more in number and denser than those in the control group. The average mineral apposition rate of the bone in the BG group was also higher than that in the other 2 groups (P0.05). The mechanical properties in the BG group were also significantly higher than that in the control group. Six months after implantation, similar results except mineral apposition rate can be obtained among different groups.BG can significantly improve bone microstructure of the interface in osteoporosis condition and increase the hold strength of the pedicle screw.

Target localization is one of the most important and fundamental issues in wireless sensor networks (WSNs). Because of its special topology structure, grid-based wireless sensor networks have a great advantage in target localization. In this paper, firstly, we analyze the uncertain area founded by the topology structure of single-level grid-deployed networks. A double-level grid-based localization method is proposed, which is theoretical and practical. We consider the computational complexity of the algorithm and define its localization error in both single-level and double-level grid-deployed networks. Simulation results reveal the general rule between different factors and their influences on location accuracy, which is meaningful to the real application of our method. The results prove that double-level grid-based localization method can not only keep location precision to a certain extent, but is also real-time and practical.

The purpose of this study is to set up a learning environment which is based on the multiple intelligences (MI) theory applied to the Internet. We try to find out the best suitable design style Web title for the students. Students of this experiment are divided into two groups of learning environments, i.e. the multiple intelligences learning Web title and general learning Web title. The student's cognitive style and multiple intelligences are identified by the questionnaires. The ACS test and SPSS are needed as research tools, to analyze the data obtained respectively. After analyzed the all factors, we find out the MI Web-based learning has the very positive effect on overall students especially on: field-independent students, and who just only have the naturalist or not only have one more intelligences simultaneously shown better achievements in the MI learning environment.

Zinc oxide (ZnO) has a wide band gap, high stability and a high thermal operating range that makes it a suitable material as a semiconductor for fabricating light emitting diodes (LEDs) and laser diodes, photodiodes, power diodes and other semiconductor devices. Recently, a new crystal growth for producing ZnO crystal boules was developed, which was physical vapor transport (PVT), at temperatures exceeding 1500 °C under a certain system pressure. ZnO crystal wafers in sizes up to 50 mm in diameter were produced. The conditions of ZnO crystal growth, growth rate and the quality of ZnO crystal were analyzed. Results from crystal growth and material characterization are presented and discussed. Our research results suggest that the novel crystal growth technique is a viable production technique for producing ZnO crystals and substrates for semiconductor device applications.

In this study, we illustrate the effective medium theories in the designs of three-dimensional composite metamaterials of both negative permittivity and negative permeability. The proposed metamaterial consists of random coated spheres with sizes smaller compared to the wavelength embedded in a dielectric host. Simple design rules and formulas following the effective medium models are numerically and analytically presented. We demonstrate that the revised Maxwell—Garnett effective medium theory enables us to design three-dimensional composite metamaterials through the assembly of coated spheres which are random and much smaller than the wavelength of the light. The proposed approach allows for the precise control of the permittivity and the permeability and guides a facile, flexible, and versatile way for the fabrication of composite metamaterials.

In this paper, a 4H-SiC semi-superjunction (SJ) Schottky barrier diode is analysed and simulated. The semi-SJ structure has an optimized design and a specific on-resistance lower than that of conventional SJ structures, which can be achieved without increasing the process difficulty. The simulation results show that the specific on-resistance and the softness factor depend on the aspect and thickness ratios, and that by using the semi-SJ structure, specific on-resistance can be reduced without decreasing the softness factor. It is observed that a trade-off exists between the specific on-resistance and the softness of the diode.

A high oxidation state alkylnitridoosmium complex, [Os(N)(CH2SiMe3)4][NBun4] acts as a nucleophile in reactions with alkyl halides. Alkylimido complexes, Os(NR)(CH2SiMe3)4, are produced. The reaction between [Os(N)(CH2SiMe3)4]− [NBun4] and MeI is second order with k2= 9.5 x 10−5 sect1 M−1 at 23°C in CD2Cl2 under pseudo first order conditions. The entropy of activation, ΔS≠, was found to be −10.6 ± 0.5 cal M−1 K−1 and the enthalpy of activation, ΔH≠, was found to be 19.6 ± 0.2 kcal M−1. The reaction proceeds faster in polar, non-coordinating solvents than in either non-polar solvents or in solvents which can coordinate to the osmium center.