Exogenous dsRNA Induces RNA Interference of a Chalcone Synthase Gene in Arabidopsis thaliana

Recent investigations have shown the possibility of artificial induction of RNA interference (RNAi) via plant foliar treatments with naked double-stranded RNA (dsRNA) to silence essential genes in plant fungal pathogens or to target viral RNAs. Furthermore, several studies have documented the downregulation of plant endogenous genes via external application of naked gene-specific dsRNAs and siRNAs to the plant surfaces. However, there are limited studies on the dsRNA processing and gene silencing mechanisms after external dsRNA application. Such studies would assist in the development of innovative tools for crop improvement and plant functional studies. In this study, we used exogenous gene-specific dsRNA to downregulate the gene of chalcone synthase (CHS), the key enzyme in the flavonoid/anthocyanin biosynthesis pathway, in Arabidopsis. The nonspecific NPTII-dsRNA encoding the nonrelated neomycin phosphotransferase II bacterial gene was used to treat plants in order to verify that any observed effects and processing of AtCHS mRNA were sequence specific. Using high-throughput small RNA (sRNA) sequencing, we obtained six sRNA-seq libraries for plants treated with water, AtCHS-dsRNA, or NPTII-dsRNA. After plant foliar treatments, we detected the emergence of a large number of AtCHS- and NPTII-encoding sRNAs, while there were no such sRNAs after control water treatment. Thus, the exogenous AtCHS-dsRNAs were processed into siRNAs and induced RNAi-mediated AtCHS gene silencing. The analysis showed that gene-specific sRNAs mapped to the AtCHS and NPTII genes unevenly with peak read counts at particular positions, involving primarily the sense strand, and documented a gradual decrease in read counts from 17-nt to 30-nt sRNAs. Results of the present study highlight a significant potential of exogenous dsRNAs as a promising strategy to induce RNAi-based downregulation of plant gene targets for plant management and gene functional studies.     

核桃花中绿原酸、新绿原酸和隐绿原酸的提取工艺响应面优化及其抗氧化活性研究

目的 探讨干燥核桃花中绿原酸、隐绿原酸和新绿原酸的最佳提取工艺及其抗氧化活性。方法 采用超声波辅助乙醇法提取核桃花中的绿原酸、新绿原酸和隐绿原酸, 以提取温度、乙醇体积分数、提取时间和料液比为自变量, 以得率的总评归一值(overall desirability, OD)为因变量, 应用响应面对提取工艺进行优化, 并通过体外实验评价了其抗氧化活性。结果 最佳提取工艺为：温度77℃, 乙醇体积分数67%, 提取时间28 min, 料液比1: 23(g/mL)。此时,绿原酸、新绿原酸和隐绿原酸的总提取率最佳, OD值可达0.91, 对DPPH自由基、ABTS自由基和羟自由基的半数抑制浓度(Half-maximal Inhibitory concentrations, IC50)值分别为41.63、41.68和34.43 μg/mL。结论 优选出的工艺稳定可行, 且提取液具有较好的体外抗氧化活性, 为核桃花中绿原酸类成分的进一步开发利用提供了理论基础。     

江南园林文化因子提取及设计应用研究

目的构建江南园林文化因子的提取模型,科学而理性地挖掘与提炼具有识别性的设计元素,结合实践以验证方法的可行性和可靠性。方法建立江南园林文化因子提取及应用研究框架,采用定性和定量相结合的方法提取花窗、洞门、色彩、意蕴4个设计因子,计算各因子的重要性并将研究成果应用于产品设计。结论使用科学的因子提取方法能够提取特征鲜明的江南园林文化设计因子,提升产品的文化内涵,以系统、科学、合理的方法指导创意产品设计实践。     

天然气超声流量计健康状态评价方法初探

目的天然气计量用超声流量计性能诊断主要采用设定参数阈值的诊断方法,为降低误报、漏报等情况,建立了基于生成对抗神经网络(generated adjoint neural network,GAN)和高维非线性无监督学习的超声流量计健康状态系统诊断方法。 方法采用GAN对原始数据进行学习、生成和扩充,保障超声流量计健康状态诊断建模的数据基础,提取超声流量计在运行过程中的健康状态参数并对其进行时序分析,采用高维非线性无监督聚类学习方法,结合超声流量计失效模式分析,对超声流量计设备进行在线的健康状态诊断。 结果结合超声流量计工作原理和现场实际采集数据,对生成的故障数据集进行了验证。 结论能够准确识别超声流量计当前状态,显著解决传统阈值法误报率、漏报率高的问题,为超声流量计健康诊断的统一管理与开发给予支撑。      

基于神经网络的图像来源识别方法比较研究

随着手机等便携式智能电子设备的普及,图像已成为最重要的信息载体之一,在新闻、社交及司法等领域发挥着重要作用.在享用电子图像带来便捷性的同时,图像处理工具给不法分子通过篡改电子图像实施诈骗等犯罪活动提供了可能,识别图像来源、辨别图像真伪已成为遏制和惩罚此类犯罪活动的重要技术手段.该文讨论了神经网络在图像源识别中的应用方法...     

Unraveling Membrane Perturbations Caused by the Bacterial Riboregulator Hfq

Hfq is a pleiotropic regulator that mediates several aspects of bacterial RNA metabolism. The protein notably regulates translation efficiency and RNA decay in Gram-negative bacteria, usually via its interaction with small regulatory RNAs. Previously, we showed that the Hfq C-terminal region forms an amyloid-like structure and that these fibrils interact with membranes. The immediate consequence of this interaction is a disruption of the membrane, but the effect on Hfq structure was unknown. To investigate details of the mechanism of interaction, the present work uses different in vitro biophysical approaches. We show that the Hfq C-terminal region influences membrane integrity and, conversely, that the membrane specifically affects the amyloid assembly. The reported effect of this bacterial master regulator on membrane integrity is discussed in light of the possible consequence on small regulatory RNA-based regulation.     

匹配追踪消噪在故障诊断中的应用

为了更有效地消除齿轮泵振动信号的噪声,引入匹配追踪消噪方法。该方法采用匹配追踪算法,通过迭代提取信号的相关结构,再利用提取的相关结构来估计信号的逼近,从而达到消噪的目的。在此基础上,采用中值滤波和线性回归算法来终止迭代过程。最后利用该方法对齿轮泵壳体的振动信号进行消噪处理。结果表明,该方法保留了信号的基本特征,有效地消除了大量的噪声。     

融合对抗训练及全局指针的实体关系联合抽取

实体关系联合抽取旨在从文本中抽取出实体关系三元组, 是构建知识图谱十分重要的步骤之一. 针对实体关系抽取中存在的信息表达能力不强、泛化能力较差、实体重叠和关系冗余等问题, 提出了一种实体关系联合抽取模型RGPNRE. 使用RoBERTa预训练模型作为编码器, 提高了模型的表达信息能力. 在训练过程中引入了对抗训练, 提升了模型的泛化能力. 使用全局指针, 解决了实体重叠的问题. 使用关系预测, 排除不可能的关系, 减少了冗余的关系. 在基于schema的中文医学信息抽取数据集CMeIE上进行的实体关系抽取实验表明, 模型的F1值比基准模型提升了约2个百分点, 在实体对重叠的情况下, 模型的F1值提升了近10个百分点, 在单一实体重叠情况下, 模型的F1值提升了大约1个百分点, 说明该模型能够更准确地提取实体关系三元组, 从而有效提升知识图谱构建的准确度. 在含有1–5个三元组的对比实验中, 在拥有4个三元组的句子中, 模型的F1值提升了约2个百分点, 而在拥有5个及以上三元组的复杂句子中, F1值提升了约1个百分点, 说明该模型能够较好地处理复杂句子场景.     

跨层协同注意和通道分组注意的细粒度图像分类

细粒度图像分类的主要挑战在于类间的高度相似性和类内的差异性. 现有的研究多数基于深层的特征而忽略了浅层细节信息, 然而深层的语义特征由于多次卷积和池化操作往往会丢失大量的细节信息. 为了更好地整合浅层和深层的信息, 提出了基于跨层协同注意和通道分组注意的细粒度图像分类方法. 首先, 通过ResNet50加载预训练模型作为骨干网络提取特征, 由最后3个阶段提取的特征以3个分支的形式输出, 每一个分支的特征通过跨层的方式与其余两个分支的特征计算协同注意并交互融合, 其中最后一个阶段的特征经过通道分组注意模块以增强语义特征的学习能力. 模型训练可以高效地以端到端的方式在没有边界框和注释的情况下进行训练, 实验结果表明, 该算法在3个常用细粒度图像数据集CUB-200-2011、Stanford Cars和FGVC-Aircraft上的准确率分别达到了89.5%、94.8%和94.7%.     

Next-Generation Grade and Survival Expression Biomarkers of Human Gliomas Based on Algorithmically Reconstructed Molecular Pathways

In gliomas, expression of certain marker genes is strongly associated with survival and tumor type and often exceeds histological assessments. Using a human interactome model, we algorithmically reconstructed 7494 new-type molecular pathways that are centered each on an individual protein. Each single-gene expression and gene-centric pathway activation was tested as a survival and tumor grade biomarker in gliomas and their diagnostic subgroups (IDH mutant or wild type, IDH mutant with 1p/19q co-deletion, MGMT promoter methylated or unmethylated), including the three major molecular subtypes of glioblastoma (proneural, mesenchymal, classical). We used three datasets from The Cancer Genome Atlas and the Chinese Glioma Genome Atlas, which in total include 527 glioblastoma and 1097 low grade glioma profiles. We identified 2724 such gene and 2418 pathway survival biomarkers out of total 17,717 genes and 7494 pathways analyzed. We then assessed tumor grade and molecular subtype biomarkers and with the threshold of AUC > 0.7 identified 1322/982 gene biomarkers and 472/537 pathway biomarkers. This suggests roughly two times greater efficacy of the reconstructed pathway approach compared to gene biomarkers. Thus, we conclude that activation levels of algorithmically reconstructed gene-centric pathways are a potent class of new-generation diagnostic and prognostic biomarkers for gliomas.