UCL Discovery
UCL home » Library Services » Electronic resources » UCL Discovery

Computational analysis of protein-carbohydrate interactions

Taroni, Chiara; (1998) Computational analysis of protein-carbohydrate interactions. Doctoral thesis (Ph.D), UCL (University College London). Green open access

[thumbnail of Computational_analysis_of_prot.pdf] Text

Download (12MB)


The subject of this work is the structural and functional study of protein-carbohydrate complexes. The goal is to identify general features characterising protein-carbohydrate interactions and use the acquired knowledge for the prediction of sugar binding-sites on proteins. As a first step, a program was developed computing pyranose ring conformations of sugars that bind to proteins. The comparison of sugar conformations in bound and free monosaccharides showed that sugars can be slightly distorted on binding, but still adopt strainless conformations. Secondly, a set of non homologous carbohydrate-binding proteins was selected on the basis of sequence and structural alignment. Two main classes of complexes have been identified; those in which the ligand is bound in a deep cleft and is considerably buried by the protein (typically enzymes and periplasmic sugar binding proteins), and those in which the carbohydrate binds in a shallow cleft and is left rather exposed to the solvent(mainly lectins and immunoglobulins). Networks of hydrogen bonds, involving principally planar, polar or charged side chains, are an essential feature of protein-carbohydrate complexes. Solvent molecules frequently mediate hydrogen bonds. The stacking of aromatic side chains on sugar rings, also appears to be crucial in the formation of protein-carbohydrate complexes. Tools were developed for determining the amino acid composition of sugar binding sites which revealed that certain amino acids show a strong propensity for being in contact with the sugar ligands. Finally an algorithm developed to compare properties of surface patches on a protein and predict regions with protein-protein interface characteristics, was modified to consider protein-sugar interfaces. Six parameters were explored to determine distinguishing properties of carbohydrate binding-sites. Three of the parameters were found to discriminate the observed sugar binding sites from the remaining protein surface and were subsequently used to calculate the probability of a surface patch being carbohydrate binding site. The overall accuracy of prediction achieved is 64%. Predictions were very good for enzymes (91% success), less effective for lectins (43% success).

Type: Thesis (Doctoral)
Qualification: Ph.D
Title: Computational analysis of protein-carbohydrate interactions
Open access status: An open access version is available from UCL Discovery
Language: English
Additional information: Thesis digitised by ProQuest.
Keywords: Biological sciences; Sugar binding sites
URI: https://discovery.ucl.ac.uk/id/eprint/10101942
Downloads since deposit
Download activity - last month
Download activity - last 12 months
Downloads by country - last 12 months

Archive Staff Only

View Item View Item