Accurate prediction of protein dihedral angles through conditional random field

Abstract

Identifying local conformational changes induced by subtle differences on amino acid sequences is critical in exploring the functional variations of the proteins. In this study, we designed a computational scheme to predict the dihedral angle variations for different amino acid sequences by using conditional random field. This computational tool achieved an accuracy of 87% and 84% in 10-fold cross validation in a large data set for ϕ and ψ, respectively. The prediction accuracies of ϕ and ψ are positively correlated to each other for most of the 20 types of amino acids. Helical amino acids can achieve higher prediction accuracy in general, while amino acids in beet sheet have higher accuracy at specific angular regions. The prediction accuracy of ϕ is negatively correlated with amino acid flexibility represented by Vihinen Index. The prediction accuracy of ϕ can also be negatively correlated with angle distribution dispersion.

Keywords

conditional random field / flexibility / angle distribution dispersion

References

[1] Ahmadi Adl A, Nowzari-Dalini A, Xue B, Uversky V N, Qian X (2012). Accurate prediction of protein structural classes using functional domains and predicted secondary structure sequences. J Biomol Struct Dyn, 29(6): 623–633

[2] Ashkenazi A, Presta L G, Marsters S A, Camerato T R, Rosenthal K A, Fendly B M, Capon D J (1990). Mapping the CD4 binding site for human immunodeficiency virus by alanine-scanning mutagenesis. Proc Natl Acad Sci USA, 87(18): 7150–7154

[3] Chang D T, Huang H Y, Syu Y T, Wu C P (2008). Real value prediction of protein solvent accessibility using enhanced PSSM features. BMC Bioinformatics, 9(Suppl 12): S12

[4] Cunningham B C, Wells J A (1989). High-resolution epitope mapping of hGH-receptor interactions by alanine-scanning mutagenesis. Science, 244(4908): 1081–1085

[5] Dang T H, Van Leemput K, Verschoren A, Laukens K (2008). Prediction of kinase-specific phosphorylation sites using conditional random fields. Bioinformatics, 24(24): 2857–2864

[6] Ellegren H (2008). Comparative genomics and the study of evolution by natural selection. Mol Ecol, 17(21): 4586–4596

[7] Faraggi E, Xue B, Zhou Y (2009). Improving the prediction accuracy of residue solvent accessibility and real-value backbone torsion angles of proteins by guided-learning through a two-layer neural network. Proteins, 74(4): 847–856

[8] Faraggi E, Yang Y, Zhang S, Zhou Y (2009). Predicting continuous local structure and the effect of its substitution for secondary structure in fragment-free protein structure prediction. Structure, 17(11): 1515–1527

[9] Gibbs C S, Zoller M J (1991). Identification of electrostatic interactions that determine the phosphorylation site specificity of the cAMP-dependent protein kinase. Biochemistry, 30(22): 5329–5334

[10] Green J R, Korenberg M J, Aboul-Magd M O (2009). PCI-SS: MISO dynamic nonlinear protein secondary structure prediction. BMC Bioinformatics, 10(1): 222

[11] Helles G, Fonseca R (2009). Predicting dihedral angle probability distributions for protein coil residues from primary sequence using neural networks. BMC Bioinformatics, 10(1): 338

[12] Lafferty J, McCallum A, Pereira F(2001) Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data. Proceedings of the 18th International Conference on Machine Learning 2001 (ICML 2001), pages 282–289.

[13] Liu Y, Carbonell J, Weigele P, Gopalakrishnan V (2006). Protein fold recognition using segmentation conditional random fields (SCRFs). J Comput Biol, 13(2): 394–406

[14] Morrison K L, Weiss G A (2001). Combinatorial alanine-scanning. Curr Opin Chem Biol, 5(3): 302–307

[15] Sutton C, McCallum A (2007)An Introduction to Conditional Random Fields for Relational Learning. In: Getoor L, Taskar B, ed. Introduction to Statistical Relational Learning. MIT Press

[16] Tokuriki N, Tawfik D S (2009). Stability effects of mutations and protein evolvability. Curr Opin Struct Biol, 19(5): 596–604

[17] Vihinen M, Torkkila E, Riikonen P (1994). Accuracy of protein flexibility predictions. Proteins, 19(2): 141–149

[18] Wallach H (2004). Conditional Random Fields: An Introduction. University of Pennsylvania CIS Technical Report MS-CIS-04-21

[19] Wang G, Dunbrack R L Jr (2003). PISCES: a protein sequence culling server. Bioinformatics, 19(12): 1589–1591

[20] Wang L, Sauer U H (2008). OnD-CRF: predicting order and disorder in proteins using [corrected] conditional random fields. Bioinformatics, 24(11): 1401–1402

[21] Xue B, Dor O, Faraggi E, Zhou Y (2008). Real-value prediction of backbone torsion angles. Proteins, 72(1): 427–433

[22] Xue B, Faraggi E, Zhou Y (2009). Predicting residue-residue contact maps by a two-layer, integrated neural-network method. Proteins, 76(1): 176–183

[23] Zhao F, Li S, Sterner B W, Xu J (2008). Discriminative learning for protein conformation sampling. Proteins, 73(1): 228–240

[24] Zhao F, Peng J, Xu J (2010). Fragment-free approach to protein folding using conditional neural fields. Bioinformatics, 26(12): i310–i317

RIGHTS & PERMISSIONS

Higher Education Press and Springer-Verlag Berlin Heidelberg