Prediction of structural classes for protein sequences and domains—Impact of prediction algorithms, sequence representation and homology, and test procedures on accuracy

Kurgan, Lukasz; Homaeian, Leila

doi:10.1016/j.patcog.2006.02.014

Cited by 152 publications

(145 citation statements)

References 60 publications

(124 reference statements)

Supporting

Mentioning

140

Contrasting

Order By: Relevance

“…34 However, compared with the results of the ''359'' dataset, the overall predictive accuracy of the ''1189'' dataset is decreasing, which is consistent with that reported previously. 34,35 Kurgan and Homaeian used the SVM as the classifier, the results of the jackknife test for the ''359'' dataset and ''1189'' dataset were 97.0% and 52.3%, respectively. 34 Kanaka et al have developed a very powerful method, called StackingC ensemble, for predicting protein structural class, the overall success rate was 96.4% for the ''359'' dataset, while on the dataset of ''1189'' proteins, the overall success rate was only about 58.9%.…”

Section: Effect Of Sequence Homology On Resultssupporting

confidence: 89%

“…Meanwhile, a nonredundant dataset of 1189 protein domains is used to evaluate the performance of the proposed method. The overall predictive accuracy achieved for the jackknife test is as high as 59.2%, which is higher than those by the StackingC ensemble, 35 about 2% higher than those by FKNN classifier 39 and about 6% higher than those by Logistic regression, 34 Bayer classifier 10 and SVM. 34 However, compared with the results of the ''359'' dataset, the overall predictive accuracy of the ''1189'' dataset is decreasing, which is consistent with that reported previously.…”

Section: Effect Of Sequence Homology On Resultsmentioning

confidence: 78%

“…The overall predictive accuracy achieved for the jackknife test is as high as 59.2%, which is higher than those by the StackingC ensemble, 35 about 2% higher than those by FKNN classifier 39 and about 6% higher than those by Logistic regression, 34 Bayer classifier 10 and SVM. 34 However, compared with the results of the ''359'' dataset, the overall predictive accuracy of the ''1189'' dataset is decreasing, which is consistent with that reported previously. 34,35 Kurgan and Homaeian used the SVM as the classifier, the results of the jackknife test for the ''359'' dataset and ''1189'' dataset were 97.0% and 52.3%, respectively.…”

Section: Effect Of Sequence Homology On Resultsmentioning

confidence: 78%

See 2 more Smart Citations

Using support vector machines for prediction of protein structural classes based on discrete wavelet transform

et al. 2008

View full text Add to dashboard Cite

The prediction of secondary structure is a fundamental and important component in the analytical study of protein structure and functions. How to improve the predictive accuracy of protein structural classification by effectively incorporating the sequence-order effects is an important and challenging problem. In this study, a new method, in which the support vector machine combines with discrete wavelet transform, is developed to predict the protein structural classes. Its performance is assessed by cross-validation tests. The predicted results show that the proposed approach can remarkably improve the success rates, and might become a useful tool for predicting the other attributes of proteins as well.

show abstract

Section: Effect Of Sequence Homology On Resultssupporting

confidence: 89%

Section: Effect Of Sequence Homology On Resultsmentioning

confidence: 78%

Section: Effect Of Sequence Homology On Resultsmentioning

confidence: 78%

See 1 more Smart Citation

Using support vector machines for prediction of protein structural classes based on discrete wavelet transform

et al. 2008

View full text Add to dashboard Cite

show abstract

“…In order to test the current method strictly and facilitate the comparison, four benchmark datasets, Z277 [6], Z498 [6], 1189 [29] and 25PDB [40], were adopted as the working datasets. The former two datasets, constructed by Zhou (1998), contain 277 and 498 sequences respectively with pairwise sequence similarities of about 80%.…”

Section: Datasetsmentioning

confidence: 99%

Prediction of protein structural class based on Linear Predictive Coding of PSI-BLAST profiles

et al. 2015

View full text Add to dashboard Cite

Knowledge of protein structure plays a key role in the analysis of protein functions, protein binding, rational drug design, and many other related fields and applications. In this study, a novel feature extraction model based on linear predictive coding (LPC) and position-specific score matrices (PSSM) was proposed to predict structural class from protein sequences. First, the PSI-BLAST tool was employed to transform the original protein sequences into PSSMs. Then, the LPC, a signal processing tool, was applied to extract the features from PSSMs. The selected features were finally fed to a support vector machine to perform the prediction. Cross-validation tests on the four benchmark datasets Z277, Z498, 1189 and 25PDB, showed a significant leap in overall accuracy using the proposed method. Compared to existing methods, our method achieved better performance in prediction of protein structural class.

show abstract

“…The composition vector is a simple sequence representation which is widely used in prediction of various aspects of protein structure [30][31][32][33][34][35][36][37]. The vector is composed of the twenty amino acids, alphabetically ordered, and stores the number of occurrences of the amino acid in the sequence window (in our case the tetrapeptide).…”

Section: Composition Vectormentioning

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

Prediction of structural classes for protein sequences and domains—Impact of prediction algorithms, sequence representation and homology, and test procedures on accuracy

Cited by 152 publications

References 60 publications

Using support vector machines for prediction of protein structural classes based on discrete wavelet transform

Using support vector machines for prediction of protein structural classes based on discrete wavelet transform

Prediction of protein structural class based on Linear Predictive Coding of PSI-BLAST profiles

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

Contact Info

Product

Resources

About