山西师大郜刚生物信息学-18-2-蛋白质结构三级预测

上传人:ll****x 文档编号:243080075 上传时间:2024-09-15 格式:PPT 页数:66 大小:5.23MB
返回 下载 相关 举报
山西师大郜刚生物信息学-18-2-蛋白质结构三级预测_第1页
第1页 / 共66页
山西师大郜刚生物信息学-18-2-蛋白质结构三级预测_第2页
第2页 / 共66页
山西师大郜刚生物信息学-18-2-蛋白质结构三级预测_第3页
第3页 / 共66页
点击查看更多>>
资源描述
单击此处编辑母版标题样式,单击此处编辑母版文本样式,第二级,第三级,第四级,第五级,*,郜刚生物信息学讲义,蛋白质三级结构预测,1,蛋白质三级结构预测,2,蛋白质三维结构的预测方法,通常包括:,同源性建模和从头开始的预测方法。,对数据库中已知结构的序列的比对是预测未知序列三级结构的主要方法,也即同源建模的方法。,3,通常对于同源建模的方法过程并非统一,但基本思路是一致的,基本包括如下几个步骤:1使用未知序列作为查询来搜索已知蛋白质结构。2产生未知序列和模版序列最可能的完整比对。3以模版结构骨架作为模型,建立蛋白质骨架模型。4在靶序列有空位区域,使用环建模过程代替合适长度的片段。5给骨架模型加上侧链。6优化侧链的位置。7使用能量最小和已知的优化知识来优化结构,4,SWISS-MODEL 是一个自动化的蛋白质比较建模服务器,可以通过ExPASy服务器的web界面免费访问。,该服务器提供用户三种模式可选择: First approach mode(简捷模式),,Alignment mode(联配模式,,Project mode(项目模式),以First approach mode(简捷模式)为例,你只需要提交你的蛋白质序列,Email,name,任务名称即可(还有其他可选项,不确定时保持默认),,5,SWISS-MODEL 服务器,以First approach mode(简捷模式)为例,,只需要提交你的蛋白质序列,Email,name,任务名称即可,(还有其他可选项,不确定时保持默认),6,Swiss-model的结果出来的快,服务器会给任务生成一个id号,并给您发送邮件结果,,一般为四封,(Swiss-model新闻,欢迎词,任务纪录,结果),,最终结果将是一个PDB文件,用,Swiss-PdbViewer,打开。SWISS-MODEL还可以通过程序Swiss Pdb-Viewer访问哦,7,SWISS-MODEL同源建模法预测蛋白质三级结构,一般由5步完成:,从,待测蛋白质序列,出发,,搜索,蛋白质结构数据库得到许多相似序列(同源序列),选定其中一个作为待测蛋白质序列的,模板,;,待测蛋白质序列与选定的模板进行再次,比对,,插入各种可能的空位使两者的保守位置尽量对齐;,建模,:调整待测蛋白序列中,主链,各个原子的位置,产生与模板相同或相似的空间结构,待测蛋白质空间结构,模型,;,利用能量最小化原理,使待测蛋白质,侧链基团,处于,能量最小,的位置。,最后提供给用户三级结构。,8,9,10,但是其模板数量似乎更新的比较慢,11,12,所有结果都将通过邮件的形式反馈用户,,反馈结果的形式,由用户自己选择,如图所示:,spdbv模式,普通模式,简短模式,有人建议使用SWISS-MODEL默认的spdbv模式,这种模式可以用Swiss-PDBViewer程序打开,可以根据需要对结果进行多种操作,13,结果举例:,以如下序列为例:,NIDRPKGLAFTDVDVDSIKIAWESPQGQVSRYRVTYSSPEDGIHELFPAPDGEEDTAELQ GLRPGSEYTVSVVALHDDMESQPLIGTQSTAIPA,域值选择:0.0000000001,结果选项:Swiss-PdbViewer mode,根据前面选项不同,可能会收到多封邮件。,SWISS-MODEL反馈的第1封邮件(Welcome to SwissModel),SWISS-MODEL反馈的第2封邮件(SwissModel-Fold-Recognition),SWISS-MODEL反馈的第3封邮件(SwissModel-SecondaryStructureP.),SWISS-MODEL反馈的第4封邮件(SwissModel_TraceLog_AAAa06ZdK),SWISS-MODEL反馈的第5封邮件(SwissModel-Model-AAAa06ZdK),中,附,有可用大分子结构展示软件,14,15,另一个牛A的服务器CPHmodels,CPHmodels,/,16,另一个牛A的服务器CPHmodels,CPHmodels是采用同源建模来预测蛋白质三级结构的一个网络服务器,同时也采用了以预测距离为基础的串线(threading)算法。同源建模是先在蛋白结构数据库中搜索目标蛋白的同源蛋白作为模板,将目标蛋白与模板进行比对,再运用一定的运算方法,以模板结构为基础构建模型,进行模型优化,将构建的模型作为预测的结果。串线算法是利用折叠识别预测蛋白质结构的方法,首先从蛋白质结构数据库中挑选蛋白质结构建立折叠子数据库,以折叠子数据库中的折叠结构作为模板,将目标序列与这些模板一一匹配,通过计算打分函数的值判断匹配程度,根据打分值给模板结构排序,其中打分最高的被认为是目标序列最可能采取的折叠结构。同时该服务器运用了神经网络算法和序表(profile)序表比对,提高了结果的准确度。神经网络算法等同于能预测两个残基相互关系的若干窗口,用来预测独立的测试蛋白中的距离是否大于或小于给定的距离标准;序表是某一序列对数据库搜索后产生的序表是一个新的比对计分矩阵,该矩阵可以被粗略的看作是一个新序列,新序列含有了更多蛋白家族的信息,用新序列进行的搜索将更加有效,更有针对性地找到家族内的匹配序列。,17,另一个牛A的服务器CPHmodels,建模来预测,同时采用了以预测距离为基础的,串线(threading)算法,。,串线算法是,利用折叠识别,预测蛋白质结构的方法,,首先从蛋白质,结构数据库,中挑选蛋白质结构,建立折叠子数据库,,,以折叠子数据库中的折叠结构,作为模板,,将目标序列与这些模板一一,进行匹配,,,通过计算,打分,函数的值判断匹配程度,根据打分值给模板结构排序,其中打分最高的被认为是目标序列最可能采取的折叠结构。,18,19,20,结果,coordinates for model in,pdb format:,model.pdb,Coordinates are free for academic, government and other non-profit institutions. Commercial entities must obtain permission to use the programs encad and segmod from the GeneMine package ().,Sorry,You need a Java-capable browser to run PDB3D,PDB3D,is a high speed 3D molecule rendering applet written by Scott Legrand at UCLA.,Controls:,Left Mouse Button,: Rotate molecule,Right Mouse Button,: Scale molecule (left & right),21,详细打分注解,CPHmodels-2.0 Server - prediction results,Technical University of Denmark,-,Query sequence:,Sequence,MEMFGKIACFVLLCMVVVAPHAEALSCGQVTSGLAPCLPYLQGSGPLGGCCGGVKGLLGA,AKTPEDRKTACTCLKSAANSIKGIDTGKAAGLPGVCGVNIPYKISPSTDCSKVQ,Searching for template .,Round 0. All hits:,entry: 1T12 chain: A score: 118 E: 1e-27,entry: 1AFH chain: _ score: 89 E: 1e-18,entry: 1FK0 chain: A score: 89 E: 1e-18,entry: 1CZ2 chain: A score: 80 E: 5e-16,entry: 1CZ2 chain: A score: 80 E: 5e-16,entry: 1GH1 chain: A score: 80 E: 5e-16,Retrieving template .,Entry: 1t12,Chain: A,gunzip: stdout: Broken pipe,Making pro alignment .,Score: 188.0 bits,Identity: 74.7 %,Query: 1 ALSCGQVTSGLAPCLPYLQGSGPLGGCCGGVKGLLGAAKTPEDRKTACTCLKSAANSIKG 60,A+CGQVTS LAPCL YL+ +GPLG CCGGVK L+ +A+T EDR+ ACTCLKSAA +I G,Templ: 1 AITCGQVTSNLAPCLAYLRNTGPLGRCCGGVKALVNSARTTEDRQIACTCLKSAAGAISG 60,Query: 61 IDTGKAAGLPGVCGVNIPYKISPSTDCSKVQ 91,I+ GKAAGLP CGVNIPYKISPSTDCSKVQ,Templ: 61 INLGKAAGLPSTCGVNIPYKISPSTDCSKVQ 91,Modeling .,22,ModBase,ModBase由大名鼎鼎的Sali实验室创建和维护。,以PSI-BLAST and MODELLER为工作基础。,ModBase还包括有,可能的配体结合位点,,SNP注释,以及,蛋白质相互作用,等,23,24,MODELIRANJE,Modeller license key,25,26,申请、注册、获得密码就可开始预测了,27,28,29,30,31,2008,ggniche2008,32,33,34,35,在排队里边你可以看到当前有哪些人在申请工作,但是你的结果只有你能够看到,,也就是只有点击前面的ID号,然后通过密码输入才能看见结果。这里的服务速度,实在是太慢了,,36,37,38,3D-JIGSAW,39,40,Thank you for trying this version of 3D-JIGSAW,Your job has been put into a queue. We will e-mail you the results as soon as possibleIf you dont receive them, probably the e-mail address was wrong :-(Your job-ref is 38dc54a0Comments are appreciated.,Email,Protein name,potato_a7,Sequence,MEMFGKIACFVLLCMVVVAPHAEALSCGQV,.,Mode,automatic,41,3D-JIGSAW,界面友好,服务快捷,42,ESyPred3D的使用流程,43,ESyPred3D,- ESyPred3D is a new automated homology modeling program.,The method gets using neural networks.,by combining, weighting and screening the results of several multiple alignment programs.,The final three dimensional structure is built using the modeling package MODELLER,44,ESyPred3D的原理,1.模板选择 2.目标和模板比对步骤 3.对没有或与模板有显著不同的区域进行建模。 4.对侧链进行建模,45,46,47,48,49,SCRATCH 蛋白质三级结构预测,SCRATCH servers,SCRATCH servers,由美国加州大学开发的基于PSI-BLAST搜索程序和神经网络算法为用户提供蛋白质空间结构预测的服务器。,预测内容包括:,二级结构特性预测(包括二级结构分类预测、相对亲水性预测、残基间接触数预测)、肽链拓扑学特性预测(包括残基接触图预测、二级结构单元接触图预测)。,50,SCRATCH 蛋白质三级结构预测,SCRATCH servers,资源网址,:,51,pdb format file,How to look?,Another PPT,52,别人的一些工作结果,53,54,55,Figure 2. (a) The comparison of B-factors of molecular surfaces calculated for uncomplexed and complexed nsLTPs,with the ligand omitted for clarity. The surface color from blue to red represents b-factors from 59.29 to 14.6 A . The,GRASP program,was used to display B-factors for (a). (b)-(c) The comparison of r.m.s. differences between uncomplexed,and complexed nsLTPs. Main-chain atoms (continuous lines) and side-chain atoms (continuous rods) of each,residue are plotted in (b) and (c).,56,(a) Stereo view of a Fo-Fc omit electron density map of nsLTP: oleate showing two conformations for the,carboxylate tail. The map was contoured at the 3 s level and displayed by using the,SETOR program,.,The coordinates of this region were omitted and the protein coordinates were rened by SHELXL-9738 before the phase calculation.,(b) A,LIGPLOT,representation of the two different oleate-binding interactions.,57,Comparisons of nsLTP:myristate with human serum albumin (HSA) in complex with myristate moieties. Insight II v97.5 (Molecular Simulations, Inc.) was used for the display. The three domains of HAS bind ve myristate moieties; four are hown here, myr1,myr3, myr4, and myr5.,58,59,60,FIG. 3.,A, stereo representation of the 15 best superimposed NMR structures of rice nsLTP2 (only the backbone atoms are shown for clarity).,B, solution structure of rice nsLTP2. Disulfide bonds involved in the three-dimensional structure are shown in ball-and-stick representation.,Helix I (,green,) and helix II (,red,) are connected through a loop (Gly17Pro21). Helix II and helix III (,purple,) are joined by a sharp 90 turn. The region containing two single turn helices and the C terminus are shown in,orange,and,brown, respectively.,碱性残基的比较可以做,结构与功能相适应,有什么样的结构就有什么样的功能,有什么样的功能就需要什么样的结构,61,FIG. 5.,A, schematic representations of the cysteine pairing patterns of nsLTP1 and nsLTP2 show high similarity except for the -CXC-motif.,B, the side-chain orientations of nsLTP1 and nsLTP2 at the -C,X,C- motifs are shown with the balland-stick model. The hydrophilic Asn49 present in nsLTP1 is projected to the periphery of the protein, whereas the hydrophobic Phe36 of nsLTP2 is buried inside the molecule.,高度相似性的比较可以做,图示细微的区别在哪里,62,FIG. 6.,Stereoview of the superimposed three-dimensional structures of rice nsLTP1(,red ribbon,) and rice nsLTP2 (,blue ribbon,).,Various positions of nsLTP2 are labeled according to their primary sequences. The N-terminal backbones of the residues of nsLTP2, Val10Ala22 (1.10 ) and Ala23Gln31 (0.52 ), are closely aligned with their counterparts in nsLTP1. The residues Gln38Gly62 of nsLTP2 are aligned (4.62) with helices III and IV of nsLTP1.,立体结构的比较可以做,63,FIG. 7.,Web-Lab Viewer representation of the docking of C18-stearic acid to the hydrophobic cavities of nsLTP1 and nsLTP2,The residues covering the hydrophobic cavities (labeled in,A,) are removed in,B,to allow the fatty acid orientation to be visualized clearly. Various residues constituting the hydrophobic cavities are labeled. The fatty acid is positioned compactly in the tunnel-like hydrophobic cavity of nsLTP1, and in the case of nsLTP2 the flexible triangular hollow box accommodates the fatty acid.,CaM结合位点的比较可以做,64,FIG. 9.,Comparison of basic residue distribution of nsLTP2 with several known plant defense peptides.,A, antifungal protein from poke weed (PDB code 1dkc);,B, antifungal protein from radish (PDB code 1ayj);,C, thionin from wheat (PDB code 1gps);,D, NsLTP2.,Panels I,and,II,differ by a 180 rotation of theviewing angle.,碱性残基的比较可以做,分子表面阳电残基的分布是抗病原物的,一个特征,这一点已经被证明,I,II,A B C D,65,通过结构的模拟,功能的预测和实验的证据,(上调表达的诱导性和钙离子积累的同步性),,我们得出以下结论:,nsLTP家族在马铃薯防御青枯菌的过程中,可能是参与钙离子信号途径发挥其抗病作用的,From structural simulation,functional predication and experimental evidence,A conclusion was drew that,所以我们的图需要有:家族内结构的比较,家族外结构的比较,,阳性残基表面示意图,CaM结合位点的示意图,进化?,66,
展开阅读全文
相关资源
正为您匹配相似的精品文档
相关搜索

最新文档


当前位置:首页 > 图纸专区 > 课件教案


copyright@ 2023-2025  zhuangpeitu.com 装配图网版权所有   联系电话:18123376007

备案号:ICP2024067431-1 川公网安备51140202000466号


本站为文档C2C交易模式,即用户上传的文档直接被用户下载,本站只是中间服务平台,本站所有文档下载所得的收益归上传人(含作者)所有。装配图网仅提供信息存储空间,仅对用户上传内容的表现方式做保护处理,对上载内容本身不做任何修改或编辑。若文档所含内容侵犯了您的版权或隐私,请立即通知装配图网,我们立即给予删除!