Improved method for predicting  -turn using support vector machine

Zhang, Qidong; Yoon, Sukjoon; Welsh, William J.

doi:10.1093/bioinformatics/bti358

Cited by 54 publications

(56 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On the other hand, type VI were divided into 2 sub-types, that is, VIa [89,117] are nowadays considered as the standard (see Table 1). They are widely analyzed in molecular dynamics [118] and prediction methods have been developed [119][120][121][122][123][124][125][126]. Motifs and conformational analysis of amino acid residues adjoining β-turns in proteins have also been extensively described [127].…”

Section: Secondary Structuresmentioning

confidence: 99%

Local Protein Structures

Offmann¹,

Tyagi²,

Brevern

2007

CBIO

View full text Add to dashboard Cite

Protein structures are classically described as composed of two regular states, the α-helices and the β-strands and one non-regular and variable state, the coil. Nonetheless, this simple definition of secondary structures hides numerous limitations. In fact, the rules for secondary structure assignment are complex. Thus, numerous assignment methods based on different criteria have emerged leading to heterogeneous and diverging results. In the same way, 3 states may over-simplify the description of protein structure; 50% of all residues, i.e., the coil, are not genuinely described even when it encompass precise local protein structures.Description of local protein structures have hence focused on the elaboration of complete sets of small prototypes or "structural alphabets", able to analyze local protein structures and to approximate every part of the protein backbone. They have also been used to predict the protein backbone conformation and in ab initio / de novo methods. In this paper, we review different approaches towards the description of local structures, mainly through their description in terms of secondary structures and in terms of structural alphabets. We provide some insights into their potential applications.

show abstract

Section: Secondary Structuresmentioning

confidence: 99%

Local Protein Structures

Offmann¹,

Tyagi²,

Brevern

2007

CBIO

View full text Add to dashboard Cite

show abstract

“…SVMs have been extensively used in various machine learning problems, especially prediction, as an alternative to standard neural network approaches (Haykin 1999). Their previously successful applications include microarray analysis (Brown et al 2000), disorder prediction in proteins (Ward et al 2004), protein secondary structure prediction (Hua and Sun 2001;Zhang et al 2005), and protein solvent accessibility prediction (Nguyen and Rajapakse 2005), to name a few. The popularity of an SVM is due to its high generalization performance, its intuitive idea, its sound mathematical foundation, and its few numbers of free parameters to adjust.…”

mentioning

confidence: 99%

Evaluation of features for catalytic residue prediction in novel folds

et al. 2007

View full text Add to dashboard Cite

Structural genomics projects are determining the three-dimensional structure of proteins without full characterization of their function. A critical part of the annotation process involves appropriate knowledge representation and prediction of functionally important residue environments. We have developed a method to extract features from sequence, sequence alignments, three-dimensional structure, and structural environment conservation, and used support vector machines to annotate homologous and nonhomologous residue positions based on a specific training set of residue functions. In order to evaluate this pipeline for automated protein annotation, we applied it to the challenging problem of prediction of catalytic residues in enzymes. We also ranked the features based on their ability to discriminate catalytic from noncatalytic residues. When applying our method to a wellannotated set of protein structures, we found that top-ranked features were a measure of sequence conservation, a measure of structural conservation, a degree of uniqueness of a residue's structural environment, solvent accessibility, and residue hydrophobicity. We also found that features based on structural conservation were complementary to those based on sequence conservation and that they were capable of increasing predictor performance. Using a family nonredundant version of the ASTRAL 40 v1.65 data set, we estimated that the true catalytic residues were correctly predicted in 57.0% of the cases, with a precision of 18.5%. When testing on proteins containing novel folds not used in training, the best features were highly correlated with the training on families, thus validating the approach to nonhomologous catalytic residue prediction in general. We then applied the method to 2781 coordinate files from the structural genomics target pipeline and identified both highly ranked and highly clustered groups of predicted catalytic residues.

show abstract

“…We employed a support vector machine (SVM) classifier [42] which was previously applied to beta-turn prediction [43,44] and was shown to provide promising results in identifying beta-turn types [20]. Given a training set of data point pairs (x i , c i ), i = 1, 2, … n, where x i denotes the feature vector, c i ={-1, 1} denotes binary class label, n is the number of training data points, finding the optimal SVM is achieved by solving: where w is a vector perpendicular to wx-b=0 hyperplane that separates the two classes, C is a user defined complexity constant, i are slack variables that measure the degree of misclassification of x i for a given hyperplane, b is an offset that defines the size of a margin that separates the two classes, and z= (x) where k(x,x')= (x) (x') is a user defined kernel function.…”

Section: Svm Classifiermentioning

confidence: 99%

“…In this case, any predicted beta-turn type is considered as a generic beta-turn and the proposed method is compared against several related methods [13,22,23,27,43,44,[52][53][54] based on 7-fold cross validation on dataset 426, see Table 3. We note that several other beta-turn prediction methods, which are not included in our comparison, were also developed [55,56].…”

Section: Quality Of Beta-turn Vs Non-beta-turn Predictionsmentioning

confidence: 99%

Sequence-Only Based Prediction of β -Turn Location and Type Using Collocation of Amino Acid Pairs

Campbell¹,

Kurgan²

2008

TOBIOIJ

View full text Add to dashboard Cite

Development of accurate -turn (beta-turn) type prediction methods would contribute towards the prediction of the tertiary protein structure and would provide useful insights/inputs for the fold recognition and drug design. Only one existing sequence-only method is available for the prediction of beta-turn types (for type I and II) for the entire protein chains, while the proposed method allows for prediction of type I, II, IV, VII, and non-specific (NS) beta-turns, filling in the gap. The proposed predictor, which is based solely on protein sequence, is shown to provide similar performance to other sequence-only methods for prediction of beta-turns and beta-turn types. The main advantage of the proposed method is simplicity and interpretability of the underlying model. We developed novel sequence-based features that allow identifying beta-turns types and differentiating them from non-beta-turns. The features, which are based on tetrapeptides (entire beta-turns) rather than a window centered over the predicted residues as in the case of recent competing methods, provide a more biologically sound model. They include 12 features based on collocation of amino acid pairs, focusing on amino acids (Gly, Asp, and Asn) that are known to be predisposed to form beta-turns. At the same time, our model also includes features that are geared towards exclusion of non-beta-turns, which are based on amino acids known to be strongly detrimental to formation of beta-turns (Met, Ile, Leu, and Val).

show abstract

Improved method for predicting -turn using support vector machine

Cited by 54 publications

References 24 publications

Local Protein Structures

Local Protein Structures

Evaluation of features for catalytic residue prediction in novel folds

Sequence-Only Based Prediction of β -Turn Location and Type Using Collocation of Amino Acid Pairs

Contact Info

Product

Resources

About