Use this URL to cite or link to this record in EThOS:
Title: The molecular recognition and partner prediction for transient protein complexes : CDK-cyclin homologue interactions
Author: Quan, Xueping
Awarding Body: University of Edinburgh
Current Institution: University of Edinburgh
Date of Award: 2006
Availability of Full Text:
Access from EThOS:
Full text unavailable from EThOS. Please try the link below.
Access from Institution:
This thesis describes a novel computational strategy that combines multiple bioinformatic program components to predict specific transient protein-protein interactions. This protocol was developed focusing especially on the transient interactions between cyclin-dependent kinases (CDK) and cyclins. We adopted a comparative modelling strategy to build 3-D models of cyclins and CDKs using known human CDK and cyclin structures as templates. These modelled structures were then subjected to a large scale docking experiment with the program ZDOCK in which all cyclin-CDK combinations were considered. In the following steps of the procedure, additional selection criteria were applied to select the most compelling complexes from the ZDOCK result list. The two principal selection criteria used were the relative CDK-cyclin subunit orientation in the complex, and interface surface property correlation. Calibration of interface surface property correlation coefficients as computed by the program MOLSURFER was based on a positive reference dataset consisting of 104 true, non-homologous, transient heterodimeric protein-protein complexes, and a negative reference dataset consisting of 70 false protein-protein complexes. Prediction accuracies achieved using this approach are expected to be around 80% based on cross-validation of the interface selection criteria. The entire modelling and prediction approach has been applied to the well-characterised set of human CDKs and cyclins. Of the resulting positive predictions, 80% were in agreement with complex formation according to HPRD and Swiss-Prot annotation. Finally, when the approach was applied to 33 CDK homologues and 35 cyclin-homologues in Arabidopsis thaliana it yielded 19 mostly likely interacting CDK-cyclin pairs. The prediction strategy developed and applied in this work should be transferable to other transient heterodimer protein-protein interactions.
Supervisor: Not available Sponsor: Not available
Qualification Name: Thesis (Ph.D.) Qualification Level: Doctoral
EThOS ID:  DOI: Not available