sequence of carbohydrate in a glycoprotein? I am sorry, question does not make any sense. 

Lipids and carbohydrates are small molecules and are not proteins, therefore cannot be coded by DNA. In glycoproteins, carbohydrate sidechains are usually very heterogenous. However, you can use metal nanoparticle probes to detect the carbohydrate–protein interactions.

You can access Glycoprotein databases at http://jcggdb.jp/rcmg/gpdb/index

You will not be able to access the carbohydrate sequence there, however, you can access the glycosylation residues of glycoproteins.

Hi,

https://glycobase.nibrt.ie/glycobase/about.action

http://jcggdb.jp/rcmg/gpdb/index.action

Hope this helps.

Regards

If your protein is an enzyme, BRENDA is perhaps the major repository of information, including post-translational modifications (sometimes in old references). But to find the precise sequence of carbohydrates attached to a glycoprotein is not easy. In fact, most frequently there is no single sequence, but a series of glycoforms that simultaneously co-exist. So a search on mass spectrometry derived data is an important task.

Accordingly, it is not surprisingly that, for most glycoproteins (estimated to be 50% of the proteins in nature and 3/4 of the proteins in humans), the carbohydrate sequence of each glycan chain is not available. In fact, in addition to the presence of glycoforms, human proteins expressed in bacteria will not be N-glycosylated and when expressed in other eukarya the glycan chain may not be the same. 

One strategy (and potentially your only one) is to calculate the approximate molecular weight of the glycan chains (subtracting the molecular mass of the purified protein from the mass derived from its amino acid sequence) and observe the type of the glycan, that is, highmannose, complex or hybrid, in the case of N-linked glycans (this you will find in literature for your system). Then you may compare the typical glycan chain to the expected molecular weight - due to the presence of multiple glycoforms, is not difficult that some very similar carbohydrate sequence indeed exists. It lacks precision but, as I mentioned above, frequently is the only way to get some insight into the carbohydrate sequence attached to a given protein.

UniProt is one reference for you.

http://www.uniprot.org/help/carbohyd

Actually I am working on Herpes Simplex Virus type 1 envelop glycoprotein C.On running blast I wouldn't get identity more than 45%. Since I am not getting PDB structure of that glycoprotein I went for its modelling with I Tasser (retrived sequence from ncbi database). Still the result is not satisfactory don't know why but I guess may be due to the presence of carbohydrate molecule. If anyone can suggest me any idea then it we helpfull for me. 

Glycan chains will not interfere in your model quality. 45% identity to a template is more then enough to build a proper comparative modeling derived structure, if this identify is well distributed along the sequence. In fact, it is very frequent for crystallographers to remove glycan chains from glycoproteins before attempting crystallization. Check the protein domains, separate them and try to model each one separately.

Only move to threading and fold recognition methods if you are not able to model the protein by comparative modeling. 

Hi,

I think you should give a try with modeller

On modelling with I Tasser I got 5 PDB structures and then I plotted the Ramachandran Plot with the help of SAVES server. In that I got that the Residues in most favoured regions were 55.4%, Residues in additional allowed regions was 36.7%,

Residues in generously allowed regions 21 5.2% and Residues in disallowed regions 11 2.7%. But at the end of the page it was given that a good quality model would be expected to have over 90% in the most favoured regions. Shall I proceed with the same result or try other methods.

Try and minimize the structure that you obtained from I-tasser and then check the plotting, you will have to minimize the structure not once but few more times, or else you can try modeller and observe that structure's Ramchandran plot.

GlycoProtDB is a glycoprotein database providing information of Asn (N)-glycosylated proteins and their glycosylated site.

https://jcggdb.jp/rcmg/gpdb/index.action