OWL2Vec*: embedding of OWL ontologies

Machine Learning - Semantic embedding of knowledge graphs has been widely studied and used for prediction and statistical analysis tasks across various domains such as Natural Language Processing...     

MR-ELM: a MapReduce-based framework for large-scale ELM training in big data era

In the big data era, extreme learning machine (ELM) can be a good solution for the learning of large sample data as it has high generalization performance and fast training speed. However, the emerging big and distributed data blocks may still challenge the method as they may cause large-scale training which is hard to be finished by a common commodity machine in a limited time. In this paper, we propose a MapReduce-based distributed framework named MR-ELM to enable large-scale ELM training. Under the framework, ELM submodels are trained parallelly with the distributed data blocks on the cluster and then combined as a complete single-hidden layer feedforward neural network. Both classification and regression capabilities of MR-ELM have been theoretically proven, and its generalization performance is shown to be as high as that of the original ELM and some common ELM ensemble methods through many typical benchmarks. Compared with the original ELM and the other parallel ELM algorithms, MR-ELM is a general and scalable ELM training framework for both classification and regression and is suitable for big data learning under the cloud environment where the data are usually distributed instead of being located in one machine.     

Fabrication and electrical properties of MoS2 nanodisc-based back-gated field effect transistors

Two-dimensional (2D) molybdenum disulfide (MoS₂) is an attractive alternative semiconductor material for next-generation low-power nanoelectronic applications, due to its special structure and large bandgap. Here, we report the fabrication of large-area MoS₂ nanodiscs and their incorporation into back-gated field effect transistors (FETs) whose electrical properties we characterize. The MoS₂ nanodiscs, fabricated via chemical vapor deposition (CVD), are homogeneous and continuous, and their thickness of around 5 nm is equal to a few layers of MoS₂. In addition, we find that the MoS₂ nanodisc-based back-gated field effect transistors with nickel electrodes achieve very high performance. The transistors exhibit an on/off current ratio of up to 1.9 × 10⁵, and a maximum transconductance of up to 27 μS (5.4 μS/μm). Moreover, their mobility is as high as 368 cm²/Vs. Furthermore, the transistors have good output characteristics and can be easily modulated by the back gate. The electrical properties of the MoS₂ nanodisc transistors are better than or comparable to those values extracted from single and multilayer MoS₂ FETs.     

酶解泥鳅蛋白制备小分子肽的工艺研究

以泥鳅为原料,酶解泥鳅蛋白制备小分子肽。运用响应面分析法优化泥鳅蛋白酶解的工艺条件,探讨了泥鳅蛋白酶解过程中液料比、酶解温度、酶解时间对多肽得率的影响。结果表明,泥鳅蛋白酶解的工艺条件为:液料比1.3、酶解温度57℃、时间9.42h,肽的得率为36.3%(模型预测值为36.37%)。通过OriginPro7.0软件计算,在显著性水平P>0.05下,预测值与实际值无显著不同,表明运用响应面分析法优化泥鳅蛋白酶解工艺条件是可行的。     

Effect of denaturation during extraction on the conformational and functional properties of peanut protein isolate

Peanut protein isolate (PPI) was extracted by alkali dissolution and acid precipitation from defatted peanut flour. The effects of extraction conditions on the denaturation and functional properties of PPI were investigated. In comparison with native peanut protein (NPP) which was extracted by ammonium sulfate, the PPI extracted by alkali dissolution and acid precipitation had a higher extent of denaturation. Arachin was affected more easily by the extraction process than conarachin and led to a noticeable decrease of thermal stability of PPI. PPI contained much lower sulfhydryl and disulfide bond contents than NPP. The analyses of intrinsic fluorescence spectra indicated a more compacted tertiary conformation of NPP than PPI. Extraction process influenced the functional properties of PPI, such as protein solubility, emulsifying activity index and foaming capacity. The relatively poor functional properties of PPI might be associated with protein denaturation/unfolding and subsequent protein aggregation.

Industrial relevance

Peanut is an important oilseed crop and a well-accepted food. After oil production through thermal treatment, the defatted peanut flour is the main byproduct, which possesses a large amount of proteins. However, due to the low extraction yield and poor functional properties of these proteins, they are not well utilised in industry till now. In this work, peanut proteins were extracted by two techniques. The results indicated that extraction technique could significantly modify the functional properties of peanut proteins. Therefore, this work is helpful for industrial utilisation of peanut proteins.     

Enzymatic hydrolysis and their effects on conformational and functional properties of peanut protein isolate

The effects of limited enzymatic hydrolysis by Alcalase on the conformational and functional properties of peanut (Arachis hypogaea L.) protein isolate (PPI) were investigated. Acid subunits of arachin were most susceptible to Alcalase hydrolysis, followed by conarachin and the basic subunits of arachin. Enzymatic hydrolysis increased the thermal stability of arachin and led to a sharp increase in the number of disulphide bonds with a decrease of the sulphydryl group in PPI hydrolysates in comparison with PPI. The analysis of intrinsic fluorescence spectra indicated a more moveable tertiary conformation of PPI hydrolysates than PPI. The limited emzymatic hydrolysis improved the functional properties of PPI, such as protein solubility and gel-forming ability, but impaired the emulsifying activity index.     

无参数忆阻系统的平面偏置控制共存分岔研究

初值依赖的超级多稳定性和偏置控制的共存吸引子近年来得到学者们的广泛关注.通过在三维线性耗散系统中引入2个具有正弦忆导的忆阻元件,构建了一种新颖的、有着平面平衡点的五维忆阻动力学系统.借助理论分析和数值计算,发现忆阻初值控制的平面共存分岔现象,并揭示了忆阻初值依赖的超级多稳定性.设计了该忆阻动力学系统的模拟电路,并由PS...     

毛蚶蛋白对益生菌抗生素胁迫的保护作用

长期摄入抗生素残留的水产品可引起肠道菌群稳态失调。该研究基于恩诺沙星对四种益生菌（两歧双歧杆菌BBi32、鼠李糖乳杆菌LGG、植物乳杆菌LP45和乳双歧杆菌Probio-M8）的抑菌作用，建立了抗生素胁迫的益生菌生长模型，评价了红三鱼、黄蚬和毛蚶蛋白粗提物及其不同分离纯化组分对益生菌抗生素胁迫的保护作用。在四种益生菌中，乳双歧杆菌Probio-M8对抗生素胁迫最为敏感，三种原料中毛蚶蛋白粗提物表现出对抗生素胁迫保护作用；毛蚶蛋白粗提物经过阴离子交换柱（SepharoseFastFlow）分离得到四个组分（Fr0、Fr1、Fr2和Fr3），其分子量分别为14.3~44.3 ku（Fr0）、20.1~29.0 ku（Fr1）、14.3~97.2 ku（Fr2）和<14.3 ku（Fr3），其中Fr0可更有效地提高Probio-M8菌液的OD600从0.80提升至1.56，Fr3可将其生长代时缩短至0.98 h。毛蚶蛋白可促进乳双歧杆菌在恩诺沙星胁迫下增殖的活性，表明毛蚶蛋白具有保护益生菌免受抗生素生长胁迫、促进益生菌增殖的功能，该研究为毛蚶蛋白对肠道稳态的调节功能提供理论依据，并提出了应对体内抗生素残留的应对方案。     

Changes in the chemical composition of traditional Chinese‐type soy sauce at different stages of manufacture and its relation to taste

Chemical composition of traditional Chinese‐type soy sauce (TCSS) at different stages of manufacture (moromi fermentation and pasteurisation) were analysed. Results showed that total nitrogen, formaldehyde nitrogen, total titratable acid, reducing sugar, NaCl, water soluble peptides, free amino acids (FAAs) and Maillard reaction products in TCSS changed significantly during manufacture. Most of FAAs were released during initial stage (0–15 days) of moromi fermentation, peptides (1–5 kDa) kept dominant (69.34–80.88%) among all peptides during manufacture and Maillard reaction products increased progressively during moromi fermentation and increased sharply during pasteurisation. According to hierarchical cluster analysis and sensory evaluation, there were obvious correlation between FAAs and sensory score of over‐all in TCSS during manufacture, which indicated that balanced FAAs composition might be mainly responsible for the taste formation of TCSS. Besides, pasteurisation can further improve the over‐all taste of TCSS.     

一种离网型风力发电系统的研究

研究了一种小型离网型风力发电系统,系统采用三相不可控整流方式进行整流,使用Lc滤波电路对输出直流纹波进行滤波,采用Boost电路作为斩波稳压电路,利用PID控制反馈使输出电压稳定在600V左右。系统逆变部分采用三相五电平SPWM逆变器,可以使输出电压谐波减少,提高输出电能质量。最后对系统进行了仿真,仿真结果验证了此方法的有效性。