蛋白质二级结构预测: 基于词条的最大熵马尔科夫方法

提出了一种新的蛋白质二级结构预测方法. 该方法从氨基酸序列中提取出和自然语言中的“词”类似的与物种相关的蛋白质二级结构词条, 这些词条形成了蛋白质二级结构词典, 该词典描述了氨基酸序列和蛋白质二级结构之间的关系. 预测蛋白质二级结构的过程和自然语言中的分词和词性标注一体化的过程类似. 该方法把词条序列看成是马尔科夫链, 通过Viterbi算法搜索每个词条被标注为某种二级结构类型的最大概率, 其中使用词网格描述分词的结果, 使用最大熵马尔科夫模型计算词条的二级结构概率. 蛋白质二级结构预测的结果是最优的分词所对应的二级结构类型. 在4个物种的蛋白质序列上对这种方法进行测试, 并和PHD方法进行比较. 试验结果显示, 这种方法的Q3准确率比PHD方法高3.9%, SOV准确率比PHD方法高4.6%. 结合BLAST搜索的局部相似的序列可以进一步提高预测的准确率. 在50个CASP5目标蛋白质序列上进行测试的结果是: Q3准确率为78.9%, SOV准确率为77.1%. 基于这种方法建立了一个蛋白质二级结构预测的服务器, 可以通过http://www.insun.hit.edu.cn:81/demos/biology/index.html来访问.     

蛋白质二级结构在线服务器预测评估

蛋白质二级结构的预测,对于研究蛋白质的功能和人类生命科学意义非凡。1951年开始提出预测蛋白质二级结构,1983年对于二级结构的预测只有50%的准确率。经过多年的发展,预测方式不断的改进和完善,到如今准确率已经超过80%。但目前预测在线服务器繁多,连续自动模型评估(CAMEO)也只给出服务器三级结构的预测评估,二级结构评估还未实现。针对上述问题,选取了以下6个服务器:PSRSM、MUFOLD、SPIDER、RAPTORX、JPRED和PSIPRED,对其预测的二级结构进行评估。并且为保证测试集不在训练集内,实验数据选取蛋白质结构数据库(Protein Data Bank,PDB)最新发布的蛋白质。在基于蛋白质同源性30%、50%和70%的实验中,PSRSM取得Q3的准确率分别为91.44%、88.12%和90.17%,比其他预测服务器中最高的MUFOLD分别高出3.19%、1.33%和2.19%,证明在同一类同源性数据中PSRSM比其他服务器有更好的预测效果。除此之外实验也得到其预测的Sov准确度也比其他服务器要高。比较各类服务器的方法与结果,得出今后蛋白质二级结构预测应当重点从大数据、模板和深度学习的角度进行研究。     

蛋白质序列中的关联规则发现及其应用

随着蛋白质序列-结构分析中使用的机器学习算法越来越复杂,其结果的解释和发现过程也随之复杂化,因此有必要寻找简单且理论上可靠的方法。通过引入原理简单、理论可靠、结果具有很强实际意义的关联规则发现算法,找到了蛋白质序列中数以万计的模式。结合实例演示了如何将这些模式应用于蛋白质序列分析中,如保守区域发现、二级结构预测等。同时根据这些结果构建了一个二级结构规则库和一种简单的二级结构预测算法,实验结果表明,约81%的二级结构可以由至少一条关联规则预测得到。     

基于离散增量的支持向量机算法预测蛋白质中四类简单超二级结构

蛋白质超二级结构预测是三级结构预测的一个非常重要的中间步骤。本文从蛋白质的一级序列出发,对5793个蛋白质中的四类简单超二级结构进行预测,以位点氨基酸为参数,采用3种片段截取方式,分别用离散增量算法预测的结果不理想,将组合的离散增量值作为特征参数输入支持向量机,取得了较好的预测结果,5交叉检验的平均预测总精度达到83.0%,Matthew’s相关系数在0.71以上。     

蛋白质二级结构预测的系统误差

目前蛋白质二级结构的预测准确率徘徊在75%左右,难以作进一步提高。本文通过统计学的方法,对蛋白质的冗余数据库进行了分析。并由此证明,目前影响预测准确率继续的真正原因是蛋白质数据库本身的系统误差,系统误差大约为25%。而该误差是由于实验条件的客观原因带来的。     

蛋白质二级结构预测方法的评价

目前评价蛋白质二级结构预测方法主要考虑预测准确率,并没有充分考虑方法自身参数对方法的影响。本文提出一种新型评价方法,将内在评价与外在评价相结合评价预测方法的优劣。以基于混合并行遗传算法的蛋白质二级结构预测方法为例,通过内在评价,合理选取内在参数——切片长度和组内类别数,有效提高预测准确率,同时,通过外在评价,与其他基于随机算法的蛋白质二级结构预测算法比较和与CASP所提供的结论比较,说明了方法的有效性与正确性,以此验证内在评价和外在评价的客观性、公正性和全面性。     

蛋白质二级结构预测服务器PSRSM

蛋白质二级结构预测是蛋白质结构研究的一个重要环节,大量的新预测方法被提出的同时,也不断有新的蛋白质二级结构预测服务器出现。试验选取7种目前常用的蛋白质二级结构预测服务器:PSRSM、SPOT-1D、MUFOLD、Spider3、RaptorX,Psipred和Jpred4,对它们进行了使用方法的介绍和预测效果的评估。随机选取了PDB在2018年8月至11月份发布的180条蛋白质作为测试集,评估角度为:Q3、Sov、边界识别率、内部识别率、转角C识别率,折叠E识别率和螺旋H识别率七种角度。上述服务器180条测试数据的Q3结果分别为:89.96%、88.18%、86.74%、85.77%、83.61%,79.72%和78.29%。结果表明PSRSM的预测结果最好。180条测试集中,以同源性30%,40%,70%分类的实验结果中,PSRSM的Q3结果分别为:89.49%、90.53%、89.87%,均优于其他服务器。实验结果表明,蛋白质二级结构预测可从结合多种深度学习方法以及使用大数据训练模型方向做进一步的研究。     

自动编码器方法的蛋白质二级结构预测

蛋白质二级结构预测是进行蛋白质三级结构研究的重要基础,氨基酸的编码方式对二级结构预测有一定的影响。本文应用了一种新的组合编码方式,即将基团编码与位置特异性打分矩阵(PSSM)进行组合的编码方式。本文中提出的基团编码是针对氨基酸的一种新的编码方式,基团编码是根据氨基酸内部组成来进行编码的,由42位属性组成。本文选取位置特异性打分矩阵(PSSM)中的Blosum62进化矩阵和新的基团编码进行组合,形成新的编码方式。然后对CB513和25pdb两组数据分别进行实验。本文中将采用贝叶斯分类器与自动编码器两种方法来对这种新的编码方式进行实验,然后比较这两种方法得到的两组数据的结果。可以很明显的发现采用自动编码器的实验结果要比使用贝叶斯分类器的结果要高出1.65%。在本文的实验中,可以提取特征的自动编码器的预测准确率更好。     

用模式识别方法预测膜蛋白RC、BR和RH的二级结构

膜蛋白的结构预测在目前比较困难.本文利用已建立的模式识别方法预测了三个典型的膜蛋白RC,BR和RH的二级结构,预测结果与实验资料的符合率与该方法用于球蛋白时的结果相仿,是成功的.本文进一步完善了模式识别预测蛋白质二级结构的方法.建立了针对球蛋白二级结构预测的多分类方法,预测精度大于60%.事实证明这是一种较好的结构预测方法,鉴于目前国内外运用模式识别方法进行结构预测研究的还不多见,我们拟进一步发展完善这一方法.     

蛋白质二级结构预测方法初探

蛋白质二级结构的预测是生物信息学中一个重要的研究课题,在对蛋白质组的研究中也是最具难度的一个问题。进行二级结构预测对于理解蛋白质结构与功能的关系,以及分子设计、生物制药等领域都有重要的现实意义。同时也是一级结构与三级结构所联系的媒介,也为三级结构的研究打下基础。虽然目前预测的方法有几十种,但准确率最高的也只有70%多,本文对于目前方法进行分析,希望从中得到更加准确的方法。     

Contact patterns between helices and strands of sheet define protein folding patterns

Comparing and classifying protein folding patterns allows organizing the known structures and enumerating possible protein structural patterns including those not yet observed. We capture the essence of protein folding patterns in a concise tableau representation based on the order and contact patterns of secondary structures: helices and strands of sheet. The tableaux are intelligible to both humans and computers. They provide a database, derived from the Protein Data Bank, mineable in studies of protein architecture. Using this database, we have: (i) determined statistical properties of secondary structure contacts in an unbiased set of protein domains from ASTRAL, (ii) observed that in 98% of cases, the tableau is a faithful representation of the folding pattern as classified in SCOP, (iii) demonstrated that to a large extent the local structure of proteins indicates their complete folding topology, and (iv) studied the use of the representation for fold identification.     

Left-handed polyproline II helix formation is (very) locally driven

The left-handed polyproline II helix (PPII) is believed to be the preferred conformation for proline-rich regions of sequence in proteins. Such regions have been postulated to be protein-protein interaction domains. The formation of this structure is studied here using simple Monte Carlo computer simulations employing the hard sphere potential. It is found that polyproline sequences adopt only the PPII structure in the simulations. Non-proline, non-glycine residues inserted as guests into polyproline host peptides are conformationally restricted by the following proline residues and tend to be part of the PPII helix. It is found through insertion of two alanine residues into polyproline that the PPII structure is not propagated through more than one non-proline residue. This finding calls into question the hypothesis that proline-rich regions will preferentially adopt this structure since many such sequences are comprised of less than 50% proline residues. Proteins 33:218–226, 1998. © 1998 Wiley-Liss, Inc.     

Extended secondary structures in proteins

Supersecondary structures of proteins have been systematically searched and classified, but not enough attention has been devoted to such large edifices beyond the basic identification of secondary structures. The objective of the present study is to show that the association of secondary structures that share some of their backbone residues is a commonplace in globular proteins, and that such deeper fusion of secondary structures, namely extended secondary structures (ESSs), helps stabilize the original secondary structures and the resulting tertiary structures. For statistical purposes, a set of 163 proteins from the protein databank was randomly selected and a few specific cases are structurally analyzed and characterized in more detail. The results point that about 30% of the residues from each protein, on average, participate in ESS. Alternatively, for the specific cases considered, our results were based on the secondary structures produced after extensive Molecular Dynamics simulation of a protein–aqueous solvent system. Based on the very small width of the time distribution of the root mean squared deviations, between the ESS taken along the simulation and the ESS from the mean structure of the protein, for each ESS, we conclude that the ESSs significantly increase the conformational stability by forming very stable aggregates. The ubiquity and specificity of the ESS suggest that the role they play in the structure of proteins, including the domains formation, deserves to be thoroughly investigated.     

Patterns of protein-fold usage in eight microbial genomes: A comprehensive structural census

Eight microbial genomes are compared in terms of protein structure. Specifically, yeast, H. influenzae, M. genitalium, M. jannaschii, Synechocystis, M. pneumoniae, H. pylori, and E. coli are compared in terms of patterns of fold usage—whether a given fold occurs in a particular organism. Of the ∼340 soluble protein folds currently in the structure databank (PDB), 240 occur in at least one of the eight genomes, and 30 are shared amongst all eight. The shared folds are depleted in all-helical structure and enriched in mixed helix-sheet structure compared to the folds in the PDB. The top-10 most common of the shared 30 are enriched in superfolds, uniting many non-homologous sequence families, and are especially similar in overall architecture—eight having helices packed onto a central sheet. They are also very different from the common folds in the PBD, highlighting databank biases. Folds can be ranked in terms of expression as well as genome duplication. In yeast the top-10 most highly expressed folds are considerably different from the most highly duplicated folds. A tree can be constructed grouping genomes in terms of their shared folds. This has a remarkably similar topology to more conventional classifications, based on very different measures of relatedness. Finally, folds of membrane proteins can be analyzed through transmembrane-helix (TM) prediction. All the genomes appear to have similar usage patterns for these folds, with the occurrence of a particular fold falling off rapidly with increasing numbers of TM-elements, according to a “Zipf-like” law. This implies there are no marked preferences for proteins with particular numbers of TM-helices (e.g. 7-TM) in microbial genomes. Further information pertinent to this analysis is available at http://bioinfo.mbb.yale.edu/genome. Proteins 33:518–534, 1998. © 1998 Wiley-Liss, Inc.     

Protein loop structure prediction with flexible stem geometries

The structure prediction of loops with flexible stem residues is addressed in this article. While the secondary structure of the stem residues is assumed to be known, the geometry of the protein into which the loop must fit is considered to be unknown in our methodology. As a consequence, the compatibility of the loop with the remainder of the protein is not used as a criterion to reject loop decoys. The loop structure prediction with flexible stems is more difficult than fitting loops into a known protein structure in that a larger conformational space has to be covered. The main focus of the study is to assess the precision of loop structure prediction if no information on the protein geometry is available. The proposed approach is based on (1) dihedral angle sampling, (2) structure optimization by energy minimization with a physically based energy function, (3) clustering, and (4) a comparison of strategies for the selection of loops identified in (3). Steps (1) and (2) have similarities to previous approaches to loop structure prediction with fixed stems. Step (3) is based on a new iterative approach to clustering that is tailored for the loop structure prediction problem with flexible stems. In this new approach, clustering is not only used to identify conformers that are likely to be close to the native structure, but clustering is also employed to identify far-from-native decoys. By discarding these decoys iteratively, the overall quality of the ensemble and the loop structure prediction is improved. Step (4) provides a comparative study of criteria for loop selection based on energy, colony energy, cluster density, and a hybrid criterion introduced here. The proposed method is tested on a large set of 3215 loops from proteins in the Pdb-Select25 set and to 179 loops from proteins from the Casp6 experiment.     

Definition of a New Information-Based Per-Residue Quality Parameter

For biomolecular NMR structures typically only a poor correspondence is observed between statistics derived from the experimental input data and structural quality indicators obtained from the structure ensembles. Here, we investigate the relationship between the amount of available NMR data and structure quality. By generating datasets with a predetermined information content and evaluating the quality of the resulting structure ensembles we show that there is, in contrast to previous findings, a linear relation between the information contained in experimental data and structural quality. From this relation, a new quality parameter is derived that provides direct insight, on a per-residue basis, into the extent to which structural quality is governed by the experimental input data.     

Sequence-similar, structure-dissimilar protein pairs in the PDB

It is often assumed that in the Protein Data Bank (PDB), two proteins with similar sequences will also have similar structures. Accordingly, it has proved useful to develop subsets of the PDB from which "redundant" structures have been removed, based on a sequence-based criterion for similarity. Similarly, when predicting protein structure using homology modeling, if a template structure for modeling a target sequence is selected by sequence alone, this implicitly assumes that all sequence-similar templates are equivalent. Here, we show that this assumption is often not correct and that standard approaches to create subsets of the PDB can lead to the loss of structurally and functionally important information. We have carried out sequence-based structural superpositions and geometry-based structural alignments of a large number of protein pairs to determine the extent to which sequence similarity ensures structural similarity. We find many examples where two proteins that are similar in sequence have structures that differ significantly from one another. The source of the structural differences usually has a functional basis. The number of such proteins pairs that are identified and the magnitude of the dissimilarity depend on the approach that is used to calculate the differences; in particular sequence-based structure superpositioning will identify a larger number of structurally dissimilar pairs than geometry-based structural alignments. When two sequences can be aligned in a statistically meaningful way, sequence-based structural superpositioning provides a meaningful measure of structural differences. This approach and geometry-based structure alignments reveal somewhat different information and one or the other might be preferable in a given application. Our results suggest that in some cases, notably homology modeling, the common use of nonredundant datasets, culled from the PDB based on sequence, may mask important structural and functional information. We have established a data base of sequence-similar, structurally dissimilar protein pairs that will help address this problem (http://luna.bioc.columbia.edu/rachel/seqsimstrdiff.htm).     

Structure alignment via Delaunay tetrahedralization

A novel protein structure alignment technique has been developed reducing much of the secondary and tertiary structure to a sequential representation greatly accelerating many structural computations, including alignment. Constructed from incidence relations in the Delaunay tetrahedralization, alignments of the sequential representation describe structural similarities that cannot be expressed with rigid-body superposition and complement existing techniques minimizing root-mean-squared distance through superposition. Restricting to the largest substructure superimposable by a single rigid-body transformation determines an alignment suitable for root-mean-squared distance comparisons and visualization. Restricted alignments of a test set of histones and histone-like proteins determined superpositions nearly identical to those produced by the established structure alignment routines of DaliLite and ProSup. Alignment of three, increasingly complex proteins: ferredoxin, cytidine deaminase, and carbamoyl phosphate synthetase, to themselves, demonstrated previously identified regions of self-similarity. All-against-all similarity index comparisons performed on a test set of 45 class I and class II aminoacyl-tRNA synthetases closely reproduced the results of established distance matrix methods while requiring 1/16 the time. Principal component analysis of pairwise tetrahedral decomposition similarity of 2300 molecular dynamics snapshots of tryptophanyl-tRNA synthetase revealed discrete microstates within the trajectory consistent with experimental results. The method produces results with sufficient efficiency for large-scale multiple structure alignment and is well suited to genomic and evolutionary investigations where no geometric model of similarity is known a priori.     

蛋白质二级结构指定

蛋白质二级结构是指蛋白质骨架结构中有规律重复的构象。由蛋白质原子坐标正确地指定蛋白质二级结构是分析蛋白质结构与功能的基础,二级结构的指定对于蛋白质分类、蛋白质功能模体的发现以及理解蛋白质折叠机制有着重要的作用。并且蛋白质二级结构信息广泛应用到蛋白质分子可视化、蛋白质比对以及蛋白质结构预测中。目前有超过20种蛋白质二级结构指定方法,这些方法大体可以分为两大类:基于氢键和基于几何,不同方法指定结果之间的差异较大。由于尚没有蛋白质二级结构指定方法的综述文献,因此,本文主要介绍和总结已有蛋白质二级结构指定方法。