科创中国

目的:构建表达人LIM矿化蛋白-1(LIM mineralization protein-1,LMP-1)基因的腺病毒重组体,体外感染骨肉瘤(Osteosarcoma,OS)细胞系U2OS,并鉴定LMP-1基因在U2OS中的表达.方法:以K562细胞的cDNA为文库,采用PCR方法对LMP-1进行扩增,通过TA克隆与pGEM-T载体连接并DNA测序.双酶切后并将目的基因插入至腺病毒穿梭质粒pAdtrack-CMV,对腺病毒穿梭质粒pAdtrack-CMV-LMP-1行双酶切和DNA测序鉴定,并通过荧光显微镜、RT-qPCR和Western blot检测pAdtrack-CMV-LMP-1在HEK-293T细胞中的表达.线性化pAdtrack-CMV-LMP-1,在BJ5183菌内完成与骨架质粒pAdeasy-1的同源重组,构建重组腺病毒质粒Ad-LMP-1.通过脂质体介导,在HEK-293A细胞内包装出复制缺陷的重组腺病毒Ad-LMP-1,大量扩增、纯化并测定滴度.用Ad-LMP-1感染OS细胞系U2OS,通过荧光显微镜、RT-qPCR和Western blot检测LMP-1在U2OS细胞中的表达.结果:双酶切和DNA测序鉴定穿梭质粒pAdtrack-CMV-LMP-1构建成功,荧光显微镜证实转染了pAdtrack-CMV-LMP-1质粒的HEK-293T细胞内有绿色荧光蛋白表达,RT-qPCR和Western blot验证了LMP-1基因的表达量明显高于对照组.通过扩增、纯化,Ad-LMP-1滴度达到1.5*109pfu/ml.荧光显微镜下观察重组腺病毒感染的U2OS内有绿色荧光蛋白表达,RT-qPCR和Western blot检测发现重组腺病毒感染U2OS后,LMP-1的mRNA和蛋白表达量明显高于对照组.结论:成功构建了人LMP-1基因腺病毒重组体,为进一步的实验研究奠定了基础.

Expandable profile liner(EPL) is a promising new oil well casing cementing technique, and welding is a major EPLs connection technology. Connection of EPL is still in the stage of manual welding so far, automatic welding technology is a hotspot of EPL which is one of the key technologies to be solved. A robot for automatic welding of "8" type EPL is studied. Four quadrants of mathematical equations of the 8-shaped cross-section track of EPL, consisting of multiple arcs, are established. Mechanism program for complex cross-section welding of EPL based on angle detection is proposed according to characteristics of small size, small valleys, and large forming errors, etc. A welding velocity vector control model is established by linkage control of a welding vehicle, a small driven actuator, and a height tracking mechanism. A constant speed control model based on an angle and symmetrical analysis model of rectangular coordinate system for EPL is built. Constraint conditions of constant speed control between each section are analyzed with 4 sections in first quadrant as an example, and cooperation work mechanism of the welding vehicle and the small tracking actuator is established based on pressure detection. The constant speed control model using angle self-test can be used to avoid the need for a precise mathematical model for tracking control and to adapt manufacture and installation deviation of EPL workpiece. The model is able to solve constant speed and trajectory tracking problems of EPL cross-section welding. EPL seams welded by the studied robot are good in appearance, and non-destructive testing(NDT) shows the seams are good in quality with no welding defects. Bulge tests show that the maximum pressure of welded EPL is 35 MPa, which can fulfill expansion performance requirements.

Steering control of a capsule robot in curve environment by magnetic navigation is not yet solved completely.A petal-shaped capsule robot with less steering resistance based on multiple wedge effects is presented,and an optimization method with two processes for determining the orientation of a pre-applied universal magnetic spin vector is proposed.To realize quick and non-contact steering swimming,a fuzzy comprehensive evaluation method for optimizing the steering driving angle is presented based on two evaluation indexes including the average steering speed and the average steering trajectory deviation,achieving the initial optimal orientation of a universal magnetic spin vector.To further reduce robotic magnetic vibration,a main target method for optimizing its final orientation,which is used for fine adjustment,is employed under the constrains of the magnetic moments.Swimming experimental results in curve pipe verified the effectiveness of the optimization method,which can be effectively used to realize non-contact steering swimming of the petal-shaped robot and reduce its vibration.

详细介绍了SVPWM的原理.采用一种易于数字化实现的SVPWM算法实现三相电压型SVPWM整流器,该整流器采用dq坐标系下的双闭环控制.给出了PI调节器参数,以及主电路电感、电容的计算公式.在Matlab Simulink环境下搭建整流器仿真器模型.仿真结果表明,整流器能实现单位功率因数整流和能量双向流动,验证了SVPWM算法和控制系统设计方法的正确性.

α-乳清蛋白(α-lactalbumin,α-LA)是哺乳动物乳汁中一种重要的蛋白质,富含机体必需氨基酸和支链氨基酸,其极佳的氨基酸比例、易吸收性以及功能特异性,使得α-LA在婴幼儿个体正常生长中具有重要的意义.本实验旨在构建人α-乳清蛋白真核表达载体pIRES2-ZsGreenl-opLA,将其转染至猪(Sus scrofa)的成纤维细胞,获得稳定表达人仪一LA的细胞.根据猪基因密码子偏爱性优化并合成人α-乳清蛋白基因mRNA序列opLA,并将其定向克隆入pIRES2一ZsGreenl真核表达载体,双酶切及测序方法鉴定重组载体;脂质体介导的方法将载体转染入培养的猪成纤维细胞,荧光显微镜下观察转染效果,G418抗性筛选重组细胞;RT-PCR方法检测目的基因的表达.结果表明,通过酶切以及测序鉴定得到的重组载体pIRES2一ZsGreenl-opLA构建成功;荧光显微镜下观察,转染了plRES2-ZsGreenl-opLA和pIRES2一ZsGreenl空载体的细胞均发出绿色荧光,且荧光多集中于细胞核;筛选重组细胞的最小G418浓度为400ng/μL;RT-PCR法检测到与目的基因大小一致的片段,而未转染组和转染空载体组均未检测到.本实验成功构建真核表达载体pIRES2-ZsGreenl-opLA,并获得稳定表达目的基因的猪成纤维细胞,该细胞可作为进一步研究转基因克隆猪的供体细胞.

信息社会中在线百科已成为人们获取知识的重要途径,而在线百科的标签系统作为其重要组成部分,不仅可以帮助人们在浏览某张页面时获取其他相关页面的信息,而且对于海量文本分类,以及提高在线百科检索系统的检索效率都有很大帮助.充分利用在线百科页面间的链接关系,提出了一种基于页面间的同质性原理和向量空间模型的全新针对在线百科的标签推荐算法HVSM(homogeneous principle based vector spacemodel).该标签推荐算法具有普适性,可在不同在线百科系统间推荐标签.实验结果表明,通过与朴素推荐算法NAM(nawere commendation model)进行比较,新的推荐算法可以达到更高的准确率.并且通过对实验数据进行分析,得到了若干有益的结论,为今后的研究工作奠定了基础.

针对传统SVM主动学习中批量采样方法的不足,提出了动态可行域划分算法.从特征空间与参数空间的对偶关系入手,深入分析SVM主动学习的本质,将特征空间中对样本的标注视为参数空间中对可行域的划分;通过综合利用当前分类模型和先前标注样本两方面信息,动态地优化可行域划分方案,以确保选取的样本对模型改进的价值,最终实现更为高效的选择性采样.实验结果表明,基于动态可行域划分的SVM主动学习算法能够显著提高所选样本的信息量,从而能够在有限的标注代价下大幅提高其分类性能.