通过对蛋白质有序-无序区域分析,找到表示氨基酸的34维特征.其中,样本熵是用于计算时间序列复杂度的参数,通过比较20种氨基酸在两种区域的出现频率,将其对应为0-9的时间序列,从而计算蛋白质的复杂度.另外,使用长度为35的滑动窗口将相邻氨基酸联系起来,提高了预测准确度.最后,使用粒子群算法优化BP神经网络的节点参数,训练并实现有序-无序分类的5个网络,取均值后转化为有序-无序输出.使用DisProt数据集和R80数据集分别进行十折交叉验证,预测准确率分别达76%和87%以上.

为满足神经网络中多种位宽数据计算的动态需求,从而提升硬件资源的能效,提出一种位串行乘法器设计—以1 bit的计算逻辑为核心,将多位数据的并行乘操作转化为每个周期进行1位数据乘操作的串行计算方式.为进一步提升硬件资源的利用率,在此基础上提出多通道位串行乘法器阵列同时进行多个数据的并行计算.实验结果显示,在最大支持位宽为8 bit的条件下,单通道位串行乘法器的LUT资源使用量是并行乘法器的41%,LUT资源有效利用率是并行乘法器的1.32倍;当通道数为8时,多通道位串行乘法器阵列的LUT资源使用量是多通道并行乘法器阵列的29%.该结构实现了硬件资源和性能之间的平衡——提高硬件资源的利用率从而

提升计算效能.

设计采用双极性结型晶体管产生一个二阶温度补偿电压,并将其与一阶温度补偿电压加权叠加得到一个低温度系数的带隙电压.通过采用增大运放增益和负反馈回路提高电源抑制比.电路基于0.13 μm BCD工艺实现,使用Cadence中Spectre环境进行仿真.在工作电源电压为5 V的情况下,温度等于27 ℃时输出电压为1.209 V,电源电压抑制比为-58 dB@100 Hz,在-40-130 ℃温度范围内,输出电压变化范围为0.93 mV,平均温度系数为4.51 ppm/℃.

复制起始蛋白DnaA通过识别复制起始位点oriC上的特殊序列,并招募复制复合体起始复制,它通过绑定ATP或ADP来对复制起始的开始进行调节.DNA聚合酶III的β亚基DnaN蛋白,负责将DNA聚合酶III连接到已经解旋的DNA上,以开始复制过程.通过过表达复制起始相关蛋白DnaA和DnaN,促进了菌体的繁殖能力,优化了菌体的生长状态,缩短了菌体倍增时间,为构建Pseudomonas mendocina工程菌株提供了参考.

Glioblastoma (GBM) is a malignant tumor with a high incidence rate and high invasion in human central nervous system tumors. However, the causes of the occurrence and development of glioblastoma are not clear. Although Ginkgo Folium (GF) extract can reduce the growth of U-87 glioma cell lines in nude mice, the action mechanism remains unclear. In this study, network pharmacology, molecular docking, and cell experiment were utilized to explore the underlying molecular mechanism of GF in treating GBM. The effective ingredients and action targets of GF were obtained from the TCMSP database. STRING database was applied for protein interaction analysis and Cytoscape software was used to build the "compound-target-disease" network and PPI network. The R language was adopted as gene ontology(GO) analysis and Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway analysis. AutoDock software was used for molecular docking verification. A cell experiment was employed to verify the inhibition of GBM by the effective compounds in GF. The results showed that 27 potential effective components and 112 potential targets of GF were selected. 7 435 targets of GBM were screened from six databases, and 98 overlapping targets were obtained. In active component-target network, quercetin, kaempferol, stigmasterol, and luteolin were the important active components. In PPI network, ESR1, RELA, FOS,CCD1, EGFR, AR, HIF1A, MAPK8, NCOA1, and MDM2 were the core targets. GO and KEGG analyses showed that the main pathways of GF in treating GBM involved PI3K-Akt, MAPK, and TNF signaling pathway. The results of molecular docking verified the good binding activity of the key pharmacodynamic molecules with the core targets. Cell experiments showed that quercetin (QUE), luteolin (LUT), and kaempferol (KAE) can inhibit cell viability in GBM cell lines U251. It can be concluded GF is the result of multi-component, multi-target, and multi-pathway interaction in the treatment of GBM, which provided a theoretical basis for the clinical application of GF in the treatment of GBM.

线粒体不仅是细胞内的能量代谢中心,而且不断进行着生物发生、融合、分裂等以适应细胞内环境的变化.NRF1是核内重要的转录因子,对线粒体的生物发生过程有重要的调控作用.线粒体的功能变化与多种代谢性疾病密切相关,如肥胖、糖尿病等.高碳水化合物饮食能够引起过度肥胖及高血糖、糖尿病等症状,持续的高血糖对线粒体的状态也会有影响.在高糖诱导的小鼠肥胖模型中,持续的高血糖对棕色脂肪线粒体的影响尚不清楚.通过高糖饮食构建了肥胖小鼠模型,结果发现高糖饲喂小鼠能够显著引起脂肪的堆积,并且发现对棕色脂肪的影响最为显著.通过对线粒体关键蛋白的检测,发现持续的高血糖能够引起NRF1蛋白的表达下调,同时引起线粒体蛋白水平的显著下降.在体外培养的HeLa细胞中,也得到了相同的结果.实验表明高糖处理能够通过抑制NRF1抑制线粒体的生物发生,调控棕色脂肪对脂肪酸的代谢功能.

基于网络药理学理论,采用生物信息学手段对中药黄芪中治疗乳腺癌有效成分及其机制进行探讨.一方面通过TCMSP 中草药系统药理学平台筛选出黄芪的有效成分,另一方面通过GeneCards、OMIM、PharmGkb、TTD、DrugBank和DisGeNET等数据库查找与乳腺癌相关靶点.而后,利用筛选出的有效成分和靶点构建"药物-成分-靶点-疾病"网络图,借助 STRING 数据库平台探究蛋白质互相作用关系,并进行 GO 分析和KEGG 通路富集分析 . 采用分子对接技术将黄芪中的主要活性成分槲皮素与 JUN、RELA、ESR1、MYC、MAPK1、FOS、HIF1A、MAPK14、AKT1、CCND1和IL-2等靶点进行分子对接.最后进行槲皮素(quercetin)对人乳腺癌MCF-7和MDA-MB-435细胞模型进行体外实验验证,结果表明黄芪可能主要是通过其活性成分槲皮素实现其抗乳腺癌的作用.本研究初步探讨黄芪治疗乳腺癌具有的多靶点、多通路的潜在作用机制,可进一步为黄芪抗乳腺癌有效成分的筛选验证提供依据.

Esophageal carcinoma (ESCC) is a common type of fatal cancer with high mortality. The purpose is to identify key genes and determine the potential mechanisms involved in the development of esophageal carcinoma based on bioinformatics analysis. Chromosome maintenance 6 complex (MCM6) has been identified to be up-regulated in a variety of malignancies. However, its effects on esophageal carcinoma remain unclear. Using the GEPIA database, it was investigated how MCM6 expression varied between healthy and cancerous esophagus tissues. And the result that ESCC with high MCM6 gene expression was found. Then immunohistochemistry (IHC) assay was performed to detect MCM6 expression in both healthy and malignant esophageal tissues, and the result showed high MCM6 protein expression in ESCC. The χ2 analysis to explore the correlation between clinic-pathological features and protein expression and the tumor size of esophageal cancer was substantially associated with MCM6 expression in 78 patients in our hospital. By performing a colony formation assay, the promoting effect of MCM6 on the proliferation of esophageal cancer cells was identified. Tumor growth assay confirmed the promoting role of MCM6 in the progression of esophageal cancer in vivo. The correlation between MCM6 and PCNA was detected by database and immunohistochemistry, and we found that MCM6 was positively correlated with PCNA expression. Further, it was speculated that MCM6 might promote the proliferation of esophageal cancer cells by mediating PNCA expression. In conclusion, MCM6 might become a new target for the treatment of esophageal carcinoma.

The existence of sign-changing solution is studied for an elliptic equation involving critical Sobolev exponent and critical Hardy-Sobolev exponent. By using a compactness result, the existence of infinitely many sign-changing solutions are proved by a combination of invariant sets method and Ljusternik-Schnirelman type minimax method.

研究了一类具有可求和二分性非线性离散系统的反周期解.应用Leary-Schauder度理论,给出了非线性离散系统存在反周期解的充分条件.最后通过例子说明了主要结论在实际问题中应用.

使用加权p,q对称损失函数研究了反向帕累托分布的形状参数在刻度参数给定条件下Bayes估计的形式与性质.得到了形状参数的Bayes估计的一般形式以及在给定共轭先验下的精确形式,证明了所得Bayes估计具有可容许性以及最小最大性.最后通过模拟研究了所得估计的返真性,结果表明,基于适当的p,q对称损失函数得出Bayes估计返真性较高.

改进了磺酰脲除草剂中间体2-氨基-4-甲基-6-甲氧基-1,3,5-三嗪（均三嗪）的合成工艺.首先以甲醇、乙腈和甲基磺酸为原料，制备原乙酸三甲酯粗品，然后再与双氰胺在氯化锌催化下发生反应，通过“一锅法”制备2-氨基-4-甲基-6-甲氧基-1,3,5-三嗪粗产物，粗产物经过乙醇/乙酸乙酯重结晶后，能够以50%总产率制得高纯度的均三嗪.产品经高效液相色谱（HPLC）测定纯度为99.90%.该合成方法原料易得，分离纯化步骤简单，具有效率高、成本低、绿色环保等优点，为2-氨基-4-甲基-6-甲氧基-1,3,5-三嗪的生产提供了新的工艺路线.

通过测定菟丝草药材对15-LOX、COX-2酶的抑制率来评价其体外抗炎活性，菟丝草药材各溶剂提取物对COX-2酶有较高的抑制率，其中石油醚提取物的活性最高；而对15-LOX酶的抑制率均比较低.同时采用GC-MS技术对其脂溶性部位进行成分分析，并根据峰面积归一化法对各组分进行定量分析.通过分析，共鉴定出酯类、脂肪酸、甾类等47种成分.结果显示，菟丝草药材有一定的抗炎活性，石油醚提取物活性最强，脂溶性提取物中甾类及脂肪酸成分含量较高.

采用水热法成功合成了一个新的三元Ni(Ⅱ)基配位聚合物，{[Ni(L)(ndc)(H₂O)₂]}_n（1）（L=4,4′-双(咪唑-1-基)-联苯，H₂ndc=2,6-萘二甲酸）.配合物1为二维4-连接的sql层，具有良好的热稳定性和水稳定性.荧光传感实验表明，1可以作为高选择性和高灵敏地检测有毒Hg(Ⅱ)离子的荧光探针.

Four aminogenized nanoparticles were prepared as follows: γ-Fe₂O₃@SiO₂-cysteamine, γ-Fe₂O₃@SiO₂-APTMS, γ-Fe₂O₃@SiO₂-PEI, γ-Fe₂O₃@SiO₂-G1.0. Magnetic solid phase extraction experiments of L-phenylalanine have been carried out at a concentration of 2.5×10^-6 mol/L, different extraction times and pH gradients. The lower fluorescence intensity of supernatant indicated the better extract effects. As for γ-Fe₂O₃@SiO₂-G1.0 nanoparticles, the best adsorption was obtained at pH=5, while the adsorption time was 8 h. As for γ-Fe₂O₃@SiO₂-PEI nanoparticles, less adsorption than above could be obtained at pH=3 and 0.5 h. γ-Fe₂O₃@SiO₂-PEI nanoparticles took less time to extract L-phenylalanine. Functionalized γ-Fe₂O₃ nanoparticles are promising for solid-phase extraction of L-phenylalanine.