Combining evolutionary information and neural networks to predict protein secondary structure

Using evolutionary information contained in multiple sequence alignments as input to neural networks, secondary structure can be predicted at significantly increased accuracy. Here, we extend our previous three-level system of neural networks by using additional input information derived from multiple alignments. Using a position-specific conservation weight as part of the input increases performance. Using the number of insertions and deletions reduces the tendency for overprediction and increases overall accuracy. Addition of the global amino acid content yields a further improvement, mainly in predicting structural class. The final network system has a sustained overall accuracy of 71.6% in a multiple cross-validation test on 126 unique protein chains. A test on a new set of 124 recently solved protein structures that have no significant sequence similarity to the learning set confirms the high level of accuracy. The average cross-validated accuracy for all 250 sequence-unique chains is above 72%. Using various data sets, the method is compared to alternative prediction methods, some of which also use multiple alignments: the performance advantage of the network system is at least 6 percentage points in three-state accuracy. In addition, the network estimates secondary structure content from multiple sequence alignments about as well as circular dichroism spectroscopy on a single protein and classifies 75% of the 250 proteins correctly into one of four protein structural classes. Of particular practical importance is the definition of a position-specific reliability index. For 40% of all residues the method has a sustained three-state accuracy of 88%, as high as the overall average for homology modelling. A further strength of the method is greatly increased accuracy in predicting the placement of secondary structure segments. © 1994 Wiley-Liss, Inc.     

The secondary structure of human 28S rRNA: The structure and evolution of a mosaic rRNA gene

Summary We have determined the secondary structure of the human 28S rRNA molecule based on comparative analysis of available eukaryotic cytoplasmic and prokaryotic large-rRNA gene sequences. Examination of large-rRNA sequences of both distantly and closely related species has enabled us to derive a structure that accounts both for highly conserved sequence tracts and for previously unanalyzed variable-sequence tracts that account for the evolutionary differences in size among the large rRNAs.Human 28S rRNA is composed of two different types of sequence tracts: conserved and variable. They differ in composition, degree of conservation, and evolution. The conserved regions demonstrate a striking constancy of size and sequence. We have confirmed that the conserved regions of large-rRNA molecules are capable of forming structures that are superimposable on one another. The variable regions contain the sequences responsible for the 83% increase in size of the human large-rRNA molecule over that ofEscherichia coli. Their locations in the gene are maintained during evolution. They are G+C rich and largely nonhomologous, contain simple repetitive sequences, appear to evolve by frequent recombinational events, and are capable of forming large, stable hairpins.The secondary-structure model presented here is in close agreement with existing prokaryotic 23S rRNA secondary-structure models. The introduction of this model helps resolve differences between previously proposed prokaryotic and eukaryotic large-rRNA secondary-structure models.     

Origins of the other metazoan body plans: the evolution of larval forms

Bilaterian animal body plan origins are not solely about adult forms. Most animals have larvae with body plans, ontogenies and ecologies distinct from adults. There are two primary hypotheses for larval origins. The first hypothesis suggests that the first animals were small pelagic forms similar to modern larvae, with adult bilaterian body plans evolved subsequently. The second hypothesis suggests that adult bilaterian body plans evolved first and that larval body plans arose by interpolation of features into direct-developing ontogenies. The two hypotheses have different consequences for understanding parsimony in evolution of larvae and of developmental genetic mechanisms. If primitive metazoans were like modern larvae and distinct adult forms evolved independently, there should be little commonality of patterning genes among adult body plans. However, sharing of patterning genes is observed. If larvae arose by co-option of adult bilaterian-expressed genes into independently evolved larval forms, larvae may show morphological convergence, but with distinct patterning genes, and this is observed. Thus, comparative studies of gene expression support independent origins of larval features. Precambrian and Cambrian embryonic fossils are also consistent with direct development of the adult as being primitive, with planktonic larvae arising during the Cambrian. Larvae have continued to co-opt genes and evolve new features, allowing study of developmental evolution.     

Morphology evolution and molecular phylogeny of Pestalotiopsis (Coelomycetes) based on ITS2 secondary structure

The internal transcribed spacer 2 (ITS2) located between the 5.8S and 28S genes of the nuclear ribosomal gene cistron is conserved at the level of secondary structure rather than primary sequence. Within the fungal genus Pestalotiopsis, there were two types of ITS2 sequence patterns, and hence secondary structures, which were supported by high bootstrap values in phylogenies based on the ProfDist distance and Profile neighbor-joining algorithms. Pestalotiopsis consists of two groups that differ in color intensity of the spore as measured by optical density (OD) in three median cells of conidia comprised of five cells with one basal and two to four apical appendages. OD was quantified using a novel method with the publicly available software, Image J. OD values of species within one clade were high (dark OD >0.6), while OD values of species in the other clade were low (pale OD <0.6). However, knobbed-tipped appendages, which have been used to classify species of Pestalotiopsis, were observed in both clades. In the dark clade, knobbed-tipped appendage strains aggregated in one subclade, but in the pale group, these strains did not aggregate.     

神经网络在蛋白质二级结构预测中的应用

介绍了蛋白质二级结构预测的研究意义,讨论了用在蛋白质二级结构预测方面的神经网络设计问题,并且较详尽地评述了近些年来用神经网络方法在蛋白质二级结构预测中的主要工作进展情况,展望了蛋白质结构预测的前景。     

General combinatorics of RNA secondary structure

The total number of RNA secondary structures of a given length with minimal hairpin loop length m(m>0) and with minimal stack length l(l>0) is computed, under the assumption that all base pairs can occur. Asymptotics are derived from the determination of recurrence relations of decomposition properties.     

二级结构组件与蛋白质耐热性的关系研究

挑选了NCBI COG数据库中具有全基因组的单细胞微生物,选择其中三维结构已知的蛋白质作为研究对象,研究了不同类型的二级结构含量和长度对古细菌和细菌类蛋白质耐热性的影响作用。结果表明:耐热的古细菌类蛋白质中含有相当数量的短的3_(10)螺旋,而耐热的细菌蛋白质中含有较短的loop环。这不仅说明二级结构对蛋白质耐热性有重要的影响,还表明二级结构对古细菌和细菌类蛋白质耐热性的影响作用是不同的。     

绒泡菌目黏菌的ITS1-5.8S-ITS2二级结构的比较分析

为探讨核糖体DNA转录间隔区(r DNA ITS)的RNA二级结构在黏菌系统发育研究中的作用,以黏菌ITS通用引物PHYS4和PHYS5对绒泡菌目5属8种黏菌的r DNA ITS进行扩增和测序,利用RNA structure构建了ITS区的RNA二级结构模型。结果表明:ITS1在绒泡菌目黏菌中不能形成一个紧实的结构,但大部分物种都具有一段稳定的螺旋结构,可能对r RNA的成熟具有作用;5.8S r RNA的二级结构相似,由4个螺旋组成,主要为两种类型;基于5.8S r RNA和28S r RNA相互作用构建的ITS2的二级结构模型显示,它由一个封闭的多分支环和至少4个主要的螺旋组成,其中螺旋IV结构相对比较保守。由于ITS区的二级结构相比核苷酸序列更加保守,因此深入地分析其二级结构有助于认识其结构与进化的关系。     

Prediction of protein secondary structure from amino acid sequence

The conformational parametersP _k for each amino acid species (j=1–20) of sequential peptides in proteins are presented as the product ofP _i,k , wherei is the number of the sequential residues in thekth conformational state (k=-helix,-sheet,-turn, or unordered structure). Since the average parameter for ann-residue segment is related to the average probability of finding the segment in the kth state, it becomes a geometric mean of (P _k )_av=(P _i,k ) ^1/n with amino acid residuei increasing from 1 ton. We then used ln(P_k)_av to convert a multiplicative process to a summation, i.e., ln(P _k ) _av =(1/n)P _i,k (i=1 ton) for ease of operation. However, this is unlike the popular Chou-Fasman algorithm, which has the flaw of using the arithmetic mean for relative probabilities. The Chou-Fasman algorithm happens to be close to our calculations in many cases mainly because the difference between theirP _k and our InP _k is nearly constant for about one-half of the 20 amino acids. When stronger conformation formers and breakers exist, the difference become larger and the prediction at the N- and C-terminal-helix or-sheet could differ. If the average conformational parameters of the overlapping segments of any two states are too close for a unique solution, our calculations could lead to a different prediction.     

Chromosome rearrangements and the evolution of genome structuring and adaptability

Eukaryotes appear to evolve by micro and macro rearrangements. This is observed not only for long-term evolutionary adaptation, but also in short-term experimental evolution of yeast, Saccharomyces cerevisiae. Moreover, based on these and other experiments it has been postulated that repeat elements, retroposons for example, mediate such events. We study an evolutionary model in which genomes with retroposons and a breaking/repair mechanism are subjected to a changing environment. We show that retroposon-mediated rearrangements can be a beneficial mutational operator for short-term adaptations to a new environment. But simply having the ability of rearranging chromosomes does not imply an advantage over genomes in which only single-gene insertions and deletions occur. Instead, a structuring of the genome is needed: genes that need to be amplified (or deleted) in a new environment have to cluster. We show that genomes hosting retroposons, starting with a random order of genes, will in the long run become organized, which enables (fast) rearrangement-based adaptations to the environment. In other words, our model provides a "proof of principle" that genomes can structure themselves in order to increase the beneficial effect of chromosome rearrangements.     

Measuring the thermodynamics of RNA secondary structure formation

The thermodynamics of RNA secondary structure formation in small model systems provides a database for predicting RNA structure from sequence. Methods for making these measurements are reviewed with emphasis on optical methods and treatment of experimental errors. Analysis of experimental results in terms of simple nearest-neighbor models is presented. Some measured sequence dependences of non-Watson-Crick motifs are discussed. © 1998 John Wiley & Sons, Inc. Biopoly 44: 309–319, 1997     

Comprehensive in vivo secondary structure of the SARS-CoV-2 genome reveals novel regulatory motifs and mechanisms

1. Download : Download high-res image (143KB)
2. Download : Download full-size image

     

Novel strategies to study the role of mutation and nucleic acid structure in evolution

Natural selection processes tune genomes in the edge of the chaos imposed by mutation and drift, allowing an enduring exploration of fitter genetic networks within the constraints imposed by self-organization and the interactions of genotype and phenotype. Alternatively, evolution can be viewed from thermodynamic, kinetic or cybernetic perspectives. Regardless of insight, there is need to understand structure-function relationships at the molecular and holistic evolutionary levels. Strategies are here described that analyze genetic variation in time and trace the evolution of nucleic acid structure. Nucleic acid scanning techniques were used to measure sequence divergence and provide a direct inference of genome-wide mutation rate. This was tested for the first time in vegetatively propagating plants. The method is general and was also used in a study of mutational patterns in phytopathogenic fungi, showing there was a link between sequence and structural diversification of ribosomal gene spacers. In order to determine if this was a general phenomenon, the origin and diversification of nucleic acid secondary structure was traced using a cladistic method capable of producing rooted phylogenetic trees. Phylogenies reconstructed from primary and secondary RNA structure were congruent at all taxonomical levels, providing evidence of a strong link between phenotype and genotype favoring thermodynamic stability and dissipation of Gibbs free energy. Overall results suggest that thermodynamic principles are important driving forces of the evolutionary processes of the living world.     

The effect of population structure on the rate of evolution

Ecological factors exert a range of effects on the dynamics of the evolutionary process. A particularly marked effect comes from population structure, which can affect the probability that new mutations reach fixation. Our interest is in population structures, such as those depicted by ‘star graphs’, that amplify the effects of selection by further increasing the fixation probability of advantageous mutants and decreasing the fixation probability of disadvantageous mutants. The fact that star graphs increase the fixation probability of beneficial mutations has lead to the conclusion that evolution proceeds more rapidly in star-structured populations, compared with mixed (unstructured) populations. Here, we show that the effects of population structure on the rate of evolution are more complex and subtle than previously recognized and draw attention to the importance of fixation time. By comparing population structures that amplify selection with other population structures, both analytically and numerically, we show that evolution can slow down substantially even in populations where selection is amplified.