Skip to Main Content
Frequently Asked Questions
Submit an ETD
Global Search Box
Need Help?
Keyword Search
Participating Institutions
Advanced Search
School Logo
Files
File List
ucin1259080387.pdf (1.15 MB)
ETD Abstract Container
Abstract Header
Enhanced prediction of Phosphorylation and Disorder in Proteins
Author Info
Swaminathan, Karthikeyan
Permalink:
http://rave.ohiolink.edu/etdc/view?acc_num=ucin1259080387
Abstract Details
Year and Degree
2009, PhD, University of Cincinnati, Engineering : Biomedical Engineering.
Abstract
Over the years, many predictors of structural and functional properties of proteins have beendeveloped on the basis that this information is encoded in the protein sequence. The fact that excellent prediction techniques are available has put the spotlight on the representation of the protein sequence i.e. the input features to these techniques. In this study, we focus on the structural properties of flexible regions and assess three specific conformational flexibility parameters (i) RSA Confidence that is readily available from our in-house secondary structure and solvent accessibility predictor, SABLE (ii) X-ray structure derived B-factors, which we enhance and (iii) NMR structure derived solvent accessibility standard deviations (SASDs), which is a feature we propose here. In the case of B-factors, a combination of PSSMs and real valued SS/RSA predictions, including RSA Confidence had been used to enhance its prediction. In the case of NMR SASDs we have presented a novel predictor that exploits the same feature set as B-factors. In each case, we have developed an epsilon-support vector regression (e-SVR) model towards this. To our knowledge, the use of SASDs as input features and the development of a predictor for the same is novel. It has also been shown that it might be easier to predict SASDs as compared to B-factors. The three flexibility parameters were then applied to the prediction of conformational disorder as well as prediction of phosphorylation. In the case of disorder, we have shown through cross-validation on our training set, that the addition of RSA confidence to the input feature space, improves its prediction significantly and is further improved with the addition of B-factor and SASD predictions developed in this study. All the three parameters were then included in the final predictor, which gave the top performance on the CASP8 data set in terms of average accuracy and CASP weighted scores. Even the removal of an easy target (T0500) did not dislodge our predictor from giving the top weighted score. In the case of phosphorylation, we have shown that the addition of real-valued SS and RSA predictions significantly improve the prediction as evaluated by cross-validation on the training set and is further improved by addition of B-factors and SASDs together. Additionally, we present a comparison of one- and two-class support vector machines (SVMs) as applied to the prediction of phosphorylation. In the prediction of phosphorylation, the methods typically employ a two-class classification approach with the limitation that the set of negative examples used for training may include some sites that are simply unknown to be phosphorylated. While one-class classification techniques have been considered in the past as a solution to this problem, their performance has not been systematically compared to two-class techniques. In this study, we developed and compared one- and two-class SVM based predictors for several commonly used sets of attributes. [These predictors are being made available at http://sable.cchmc.org/] Keywords: phosphorylation, protein disorder, prediction, b-factors, solvent accessibility.
Committee
Jaroslaw Meller, PhD (Committee Chair)
Marepalli Rao, PhD (Committee Member)
Mario Medvedovic, PhD (Committee Member)
Pages
90 p.
Subject Headings
Biomedical Research
Keywords
phosphorylation
;
protein disorder
;
b-factors
;
solvent accessibility
;
prediction
Recommended Citations
Refworks
EndNote
RIS
Mendeley
Citations
Swaminathan, K. (2009).
Enhanced prediction of Phosphorylation and Disorder in Proteins
[Doctoral dissertation, University of Cincinnati]. OhioLINK Electronic Theses and Dissertations Center. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1259080387
APA Style (7th edition)
Swaminathan, Karthikeyan.
Enhanced prediction of Phosphorylation and Disorder in Proteins.
2009. University of Cincinnati, Doctoral dissertation.
OhioLINK Electronic Theses and Dissertations Center
, http://rave.ohiolink.edu/etdc/view?acc_num=ucin1259080387.
MLA Style (8th edition)
Swaminathan, Karthikeyan. "Enhanced prediction of Phosphorylation and Disorder in Proteins." Doctoral dissertation, University of Cincinnati, 2009. http://rave.ohiolink.edu/etdc/view?acc_num=ucin1259080387
Chicago Manual of Style (17th edition)
Abstract Footer
Document number:
ucin1259080387
Download Count:
752
Copyright Info
© 2009, all rights reserved.
This open access ETD is published by University of Cincinnati and OhioLINK.