Prediction Of O-Glycosylation In Proteins For Different Polypeptide Galnac-Transferases

Mucin-type O-Glycosylation is a posttranslational modification of proteins found on secreted and cell surface proteins in most animals which serves multiple important biological functions. Â In humans, mutations and changes in the expression levels of O-glycosylating polypeptide GalNac-transferases (GALNTs) have been linked to diabetes and multiple cancers. In most animals the GALNTs are compose a large family of isoforms with humans having 20 isoforms. Presently, the prediction of the sites that will be O-glycosylated is a difficult task due to each isoforms different substrate preference. In this work ISOGlyP, an isoform specific O-glycosylation prediction program, was redesigned and expanded to incorporate new programming features and to increase its overall accuracy. ISOGlyP was shown to perform as well as NetOglyc4.0, another commonly used O-glycosylation predicting program which lacks GALNT specificity. ISOGlyP was redesigned so that the core prediction program could be accommodate additional functionality, for example the inclusion of new data and the ability to perform further data analyses, in addition to providing online access at ISOGlyP.utep.edu via a web.py framework. This included the recognition of the effects of prior GalNAc glycosylation of the peptide and the differences between the glycosylation of threonine and serine residues. One novel feature that was developed the ability to generate a list of peptide sequences specific for one or more GALNTs with the exclusion of other GALNTs. To further increase the accuracy of the prediction additional features, such as protein structure, predicted protein disorder, coiled regions and relative solvent accessibility, were tested for incorporation into the predictive algorithm. Finally, an iterative approach to mimic the O-glycosylation of a protein with multiple glycosites was created. This approach will help us model how the presence of different sets of transferases could alter the overall glycosylation of a protein. It is anticipated that ISOGlyP, by linking GALNTs to the sites that they glycosylate, will be an invaluable tool for the prediction and interpretation of mucin type O-glycosylation, from which a better understand the biological roles of mucin type O-glycosylation can be obtained