Determining human plasma glycoproteins

Glikozilacija je skup čestih i složenih kotranslacijskih/postranslacijskih kovalentih modifikacija lipida i proteina. One su od velike važnosti jer utječu na funkciju i strukturu takvih kompleksnih molekula. Istraživanja kompleksnih molekula su vrlo zahtjevna i često koriste kombinacije metoda. Jedna od važnih grupa metoda koje se koriste u takvim istraživanjima su bioinformatičke metode. One se uglavnom fokusiraju na analizu podataka i simulacije. Kako raste broj radova na tu temu, raste i količina podataka koje bioinformatički alati trebaju koristiti te je nužno razvijati nove alate koji će omogućiti lakše i brže istraživanje složenih mo lekula. Ovaj diplomski rad je pokušaj razvijanja takvog alata koji olakšava pretraživanje glikoproteoma UniProt baze podataka. Razvijeni alat je testiran na 592 anotirana glikoproteina humane plazme. Računalnom analizom tih glikoproteina utvrđeno je da su oni većinom manji od 1000 aminokiselina i mase obično ispod 100 kDa. Također je utvrđeno da su njihove proteinske strukture hidrofilne i da je najčešći tip glikozilacije zabilježenih proteina humane plazme Nglikozilacija. Analiza je provedena i na peptidima proteina nastalih računalnom tripsinskom razgradnjom proteina. Tom analizom je utvrđeno da je većina takvih peptida neglikozilirana, većinom manje mase od 10 kDa, odnosno duljine 100 aminokiselina te da su manje hidrofilni od izvornih glikoproteina od kojih potječu.Glycosylation is a collection of complex cotranslational/posttranslational covalent modifications of lipids and proteins. These modifications are of great importance as they affect the function and structure of such complex molecules. Researching complex molecules is very demanding and often requires combination of different types of methods. One such group includes bioinformatic methods whose main goals are data analysis and simulations. As the number of research papers grows, the amount of data on these molecules grows as well which consequently requires better and faster tools for data analysis. This Master’s thesis was an attempt of developing such a tool to improve the ones already available for UniProt database. The developed tool was tested on a set of 592 annotated human plasma glycoproteins. Computer analysis of the aforementioned set yielded the insight that most of these proteins were smaller than 1000 amino acids with masses under 100 kDa. It has also showed that most of these glycoprotein structures are hydrophilic in nature and that by far the most common type of glycosylation is the N-glycosylation. Analysis of peptides acquried by computer trypsine digestion was conducted as well. It showed that most of such peptides are much smaller than their glycoprotein counterparts, mostly ranging in size under 10 kDa or 100 amino acids in length and are less hydrophilic than full glycoproteomic structures