We recently made large changes to the webserver infrastructure, so you might experience errors. Please report issues to health-master@dtu.dk

NetNGlyc - 1.0

N-linked glycosylation sites in human proteins

The NetNglyc server predicts N-Glycosylation sites in human proteins using artificial neural networks that examine the sequence context of Asn-Xaa-Ser/Thr sequons.

Submission

Notes: SignalP is automatically run on all sequences. A warning is displayed if a signal peptide is not detected. In transmembrane proteins, only extracellular domains may be N-glycosylated. This is currently not checked by the NetNGlyc server. Cytoplasmic and transmembrane sequence regions may be predicted to be glycosylated - this should, of course, be ignored. One transmembrane region predictor is TMHMM.

Restrictions: At most 2,000 sequences and 200,000 amino acids per submission; each sequence not more than 4,000 amino acids.
Confidentiality: The sequences are kept confidential and will be deleted after processing.

CITATIONS

For publication of results, please cite:

Gupta R, Brunak S.
Prediction of glycosylation across the human proteome and the correlation to protein function.
Pac Symp Biocomput. 2002;:310-22.
PMID: 11928486

Instructions

In order to use the NetNGlyc server for prediction on amino acid sequences:

Enter a sequence (or multiple sequences in FASTA format) in the sequence window. Alternatively, give a file name containing sequences in FASTA format (multiple sequences allowed).
The sequence must be written using the one letter amino acid code: `acdefghiklmnpqrstvwy' or `ACDEFGHIKLMNPQRSTVWY'.
Other letters will be converted to `X' and treated as unknown amino acids.
Other characters, such as whitespace and numbers, will simply be ignored.
Include Graph: A graphic illustrating glycosylation potentials across the sequence length will be generated (recommended).
Show additional thresholds: Use this option if you want the graph to include more thresholds than the default 0.5. These additional thresholds (0.32, 0.75, 0.90) are used to assign higher confidence levels for positive and negative sites. See more information in the Output Format notes.
Choose the output format: Predict only on Asparagines that occur within the Asn-Xaa-Ser/Thr triplet, or show output for all Asparagines in the sequence. Note that predictions on Asparagines that do not occur within the Asn-Xaa-Ser/Thr sequon are unlikely to be glycosylated, no matter what the prediction score. The prediction method examines sequence context beyond the Asn-Xaa-Ser/Thr sequon since both the positive and negative data sets only those Asparagines (to train on) that occur in Asn-Xaa-Ser/Thr sequons. See more information in the Output Format notes.
Press the "Submit sequence" button.
A WWW page will return the results when the prediction is ready. Response time depends on system load, but is usually only a few seconds.

Output format

# Predictions for N-Glycosylation sites in 1 sequence Name: CBG_HUMAN Length: 405
(Sequence) Asn-Xaa-Ser/Thr sequons (including Asn-Pro-Ser/Thr) are shown in blue. Asparagines predicted to be N-glycosylated are shown in red. Note that not all sequons are predicted glycosylated.
MPLLLYTCLLWLPTSGLWTVQAMDPNAAYVNMSNHHRGLASANVDFAFSLYKHLVALSPKKNIFISPVSISMALAMLSLG 80 TCGHTRAQLLQGLGFNLTERSETEIHQGFQHLHQLFAKSDTSLEMTMGNALFLDGSLELLESFSADIKHYYESEVLAMNF 160 QDWATASRQINSYVKNKTQGKIVDLFSGLDSPAILVLVNYIFFKGTWTQPFDLASTREENFYVDETTVVKVPMMLQSSTI 240 SYLHDSELPCQLVQMNYVGNGTVFFILPDKGKMNTVIAALSRDTINRWSAGLTSSQVDLYIPKVTISGVYDLGDVLEEMG 320 IADLFTNQANFSRITQDAQLKSSKVVHKAVLQLNEEGVDTAGSTGVTLNLTSKPIILRFNQPFIIMIFDHFTWSSLFLAR 400 VMNPV (Annotation line) `N' represents a predicted N-glycosylation site. `n' represents an Asn with a positive score, but not occuring within an Asn-Xaa-Ser/Thr sequon
..............................N................................................. 80 ...............N................................................................ 160 ................................................................................ 240 ...................N............................................................ 320 ................................................N............................... 400 ..... (Threshold=0.5) -------------------------------------------------------------------------------- SeqName Position Potential Jury NGlyc agreement result -------------------------------------------------------------------------------- CBG_HUMAN 31 NMSN 0.7166 (9/9) ++ <-- Predicted as N-glycosylated (++) CBG_HUMAN 96 NLTE 0.6356 (8/9) + <-- Predicted as N-glycosylated (+) CBG_HUMAN 176 NKTQ 0.3941 (7/9) - <-- A negative site CBG_HUMAN 260 NGTV 0.7400 (9/9) ++ CBG_HUMAN 330 NFSR 0.4223 (7/9) - see below for CBG_HUMAN 369 NLTS 0.6684 (9/9) ++ more information --------------------------------------------------------------------------------

Graphics in PostScript

The graph illustrates predicted N-glyc sites across the protein chain (x-axis represents protein length from N- to C-terminal). A position with a potential (vertical lines) crossing the threshold (horizontal line at 0.5) is predicted glycosylated. Additional thresholds are shown at 0.32, 0.75 and 0.90 by horizontal dotted lines. Explained below. An Encapsulated postscript format of the graph is available for including in publications.

More Notes

The Asn-Xaa-Ser/Thr sequon

N-glycosylation is known to occur on Asparagines which occur in the Asn-Xaa-Ser/Thr stretch (where Xaa is any amino acid except Proline). While this consensus tripeptide (also called the N-glycosylation sequon in many texts) may be a requirement, it is not always sufficient for the Asparagine to be glycosylated. Furthermore, there are a few known instances of N-glycosylation occuring within Asn-Xaa-Cys (a Cysteine opposed to a Serine/Threonine at the N+2 position) e.g. plasma protein C (PRTC_HUMAN), von Willebrand factor (VWF_HUMAN).

NetNGlyc attempts to distinguish glycosylated sequons from non-glycosylated ones. By default, predictions are only shown on Asn-Xaa-Ser/Thr sequons. If you choose to predict on all Asparagines, then please be careful while interpreting the output. From what we know so far, only asparagines within Asn-Xaa-Ser/Thr (and in some cases, Asn-Xaa-Cys) are N-glycosylated in vivo.

In the sequence output above, Asn-Xaa-Ser/Thr sequons are highlighted in blue, and N-glycosylated Asparagines are red. With the scores for each position, Asn-Xaa-Ser/Thr sequons can be identified (in case prediction is made on all Asparagines) by a 'SEQUON' note in the right margin.

Asn-Pro-Ser/Thr

Proline just after the Asparagine, is known to preclude N-linked glycosylation in most cases by rendering the Asparagine inaccessible. NetNGlyc has been trained to ignore this Proline position (to be able to pick up other sequence signals). Thus, Asn-Pro-Ser/Thr triplets might be predicted as glycosylated but a warning is generated. Such sites may only be worth considering if there is additional confirmatory evidence.

Thresholds and confidence

Any potential crossing the default threshold of 0.5, represents a predicted glycosylated site (as long as it occurs in the required sequon Asn-Xaa-Ser/Thr without Proline at Xaa). The 'potential' score is the averaged output of nine neural networks. For further information, the jury agreement column indicates how many of the nine networks support the prediction. The N-Glyc Result column shows one of the following outputs for predictions indicating

glycosylated sites:


   + Potential < 0.5

  ++ Potential <  0.5 AND Jury agreement (9/9)  OR Potential<0.75

 +++ Potential < 0.75 AND Jury agreement

++++ Potential < 0.90 AND Jury agreement

and non-glycosylated sites:


   - Potential < 0.5

  -- Potential < 0.5 AND Jury agreement (all nine > 0.5)

 --- Potential < 0.32 AND Jury agreement

For picking up N-glycosylation sites with high specificity (Asparagines very likely to be glycosylated), use only (++) predictions (and better) for Asparagines that occur within the Asn-Xaa-Ser/Thr triplet (no Proline at the Xaa position). Note that identifying sites this way would compromise sensitivity (you may lose some positive sites).

Warnings and notes in the right margin

SEQUON ASN-XAA-SER/THR.

If you request a prediction on all Asparagines (instead of the default to predict only on Asn-Xaa-Ser/Thr sequons), then this note will appear for Asparagine positions which do occur within the Asn-Xaa-Ser/Thr sequon.

WARNING: PRO-X1.

Proline occurs just after the Asparagine residue. This makes it highly unlikely that the Asparagine is glycosylated, presumably due to conformational constraints.

WARNING: PRO-X2.

Proline occurs at the 3rd position C-terminal to the Asparagine in question (2nd 'X' in NX[ST]X). This makes it somewhat unlikely that the Asparagine is glycosylated, but this condition is not as harsh as the PRO-X1 condition.

NetNGlyc Abstract

Contrary to widespread belief, acceptor sites for N-linked glycosylation on protein sequences, are not well characterised. The consensus sequence, Asn-Xaa-Ser/Thr (where Xaa is not Pro), is known to be a prerequisite for the modification. However, not all of these sequons are modified and it is thus not discriminatory between glycosylated and non-glycosylated asparagines. We train artificial neural networks on the surrounding sequence context, in an attempt to discriminate between acceptor and non-acceptor sequons. In a cross-validated performance, the networks could identify 86% of the glycosylated and 61% of the non-glycosylated sequons, with an overall accuracy of 76%. The method can be optimised for high specificity or high sensitivity. Apart from characterising individual proteins, the prediction method can rapidly scan complete proteomes.

Glycosylation is an important post-translational modification, and is known to influence protein folding, localisation and trafficking, protein solubility, antigenicity, biological activity and half-life, as well as cell-cell interactions. We investigate the spread of known and predicted N-glycosylation sites across functional categories of the human proteome.

CURRENT NETWORK

The network will be updated and predictions can alter due to different versions. The network is balanced to give optimal predictions whether or not you submit sequences with homology to the known N-glycosylated proteins. If however the submitted sequence is very close to or identical to the sequences in our training dataset, the accuracy can be expected to be higher than reported above.

FEEDBACK, COMMENTS AND SUGGESTIONS:

We would appreciate any confirmation or the opposite of our predictions. Since an expanded data set with additional N-glycosylated sequences would increase the performance of the network, we are very interested in receiving such material. User feedback is the only way we will learn to enhance the performance of the method. Any other comments regarding the predictions or the data may be sent to:

Ramneek Gupta

Software Downloads

Version 1.0d

Linux

Version 1.0d

Linux