Brief explanation of META PredictProtein


User information

  1. Your email Example: rost@columbia.edud
    Your entire (and entirely correct) email address (e.g.
    Note: typos will result in that no results will be returned ...
  2. One-line name of protein Example: Cytochrome C oxidase
  3. Paste, or type your sequence
    Please use only one-letter code amino acids. In particular, avoid numbers or '*', or '.'.
    For other possible input formats click here!
    Click on the button SUBMIT to request a prediction
    Click on the button CLEAR to clear all data you filled in (e.g. to restart, or to send a new request).

List of services available
various  SignalPNetOglycNetPhosNetPicoChloroP
homology modelling  SWISS-MODELCPHmodels
threading  FRSVRSAMt98
secondary structure  JPRED
transmembrane helices  TMHMMTopPredDAS

Information about services available

Various analyses

Various services for sequence analysis.
Server SignalP
Site (URL)
About The SignalP World Wide Web server predicts the presence and location of signal peptide cleavage sites in amino acid sequences from different organisms: Gram-positive prokaryotes, Gram-negative prokaryotes, and eukaryotes. The method incorporates a prediction of cleavage sites and a signal peptide/non-signal peptide prediction based on a combination of several artificial neural networks.
Options If you do not check the organism (gram-positive, gram-negative prokaryotes, or eukaryotes), predictions for all three will be returned.
Quote H Nielsen, J Engelbrecht, S Brunak, and G von Heijne: Identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Protein Engineering 10, 1-6, 1997
Authors Henrik Nielsen, Soeren Brunak (both CBS Copenhagen), and Gunnar von Heijne (Univ Stockholm, Sweden)
Contact Kristoffer Rapacki (
Server NetOglyc
Site (URL)
About The NetOglyc WWW server produces neural network predictions of mucin type GalNAc O-glycosylation sites in mammalian proteins.
  1. JE Hansen, O Lund, N Tolstrup, AA Gooley, KL Williams, and S Brunak: NetOglyc: Prediction of mucin type O-glycosylation sites based on sequence context and surface accessibility. Glycoconjugate Journal, 15, 115-130, 1998
  2. JE Hansen, O Lund, K Rapacki, and S Brunak: O-glycbase version 2.0 - A revised database of O-glycosylated proteins. Nucleic Acids Research, 25, 278-282, 1997
  3. JE Hansen, O Lund, K Rapacki J Engelbrecht, H Bohr, JO Nielsen, J-E S Hansen, and S Brunak: Prediction of O-glycosylation of mammalian proteins: Specificity patterns of UDP-GalNAc:-polypeptide N-acetylgalactosaminyltransferase. Biochemical Journal, 308, 801-813, 1995
Authors Jan Hansen (CBS, Copenhagen, Denmark)
Contact Kristoffer Rapacki (
Server NetPhos
Site (URL)
About NetPhos is a neural network-based method for predicting potential phosphorylation sites at serine, threonine or tyrosine residues in protein sequences.
Quote N Blom, S Gammeltoft, and S Brunak: Sequence- and structure-based prediction of eukaryotic protein phosphorylation Sites. J of Molecular Biology, 294, 1351-1362, 1999.
Authors Nikolaj Blom (CBS, Copenhagen, Denmark,
Contact Kristoffer Rapacki (
Server NetPico
Site (URL)
About The NetPicoRNA World Wide Web server produces neural network predictions of cleavage sites of picornaviral proteases.
Quote N Blom, J Hansen, D Blaas, and S Brunak: Cleavage site analysis in picornaviral polyproteins: Discovering cellular targets by neural networks. Protein Science, 5, 2203-2216, 1996
Authors Nikolaj Blom (CBS, Copenhagen, Denmark,
Contact Kristoffer Rapacki (
Server ChloroP
Site (URL)
About The ChloroP www-server is able to predict two things:
  • 1. cTP or no cTP
    Whether or not an amino acid sequence contains an N-terminal chloroplast transit peptide, cTP.
  • 2. Cleavage site
    The probable site for cleavage of the transit peptide (if it was predicted to exist in the first step).
  • Include the N-terminus
    It is strongly recommended to include the N-terminus of the submitted sequence. The further from the N-terminal residue the submitted sequence starts, the more difficult and unreliable will the prediction be.
  • Submit preferably 100-150 residues
    Submit if possible at least 100 and no more than 150 N-terminal residues. The lower boundary is due to the fact that the "cTP"/"no cTP" predictor was trained with input sequences of length 100 residues. However, shorter sequences may also be satisfactory predicted (it is more important that the N-terminal part is intact). The cleavage site prediction is in itself not influenced by sequence length, but restricting the submitted length to approximately 150 residues prevents the prediction of too long cTP's.
Quote O Emanuelsson, H Nielsen, and G von Heijne:ChloroP, a neural network-based method for predicting chloroplast transit peptides and their cleavage sites. Protein Science, 8, 978-984, 1999
Authors Olof Emanuelsson (CBS Copenhagen, Denmark,
Contact Kristoffer Rapacki (

Secondary structure prediction

Servers returning predictions of secondary structure based on single sequences, or sequence alignments.
Server JPRED
Site (URL)
About A consensus method for protein secondary structure prediction.
Quote J A Cuff, and G J Barton: Evaluation and improvement of multiple sequence methods for protein secondary structure prediction. PROTEINS, 34, 508-519, 1999
Authors James Cuff, and Geoff Barton (EBI Hinxton, England)
Contact James Cuff (

Prediction of transmembrane helices

Servers returning predictions of transmembrane helix location and topology based on single sequences, or sequence alignments. Note: there is no general method for detecting porin-like membrane proteins, yet.
Server TMHMM
Site (URL)
About Hidden Markov Model predicting the location of transmembrane helices and their topology. (Remark Burkhard Rost: find it to be the most original method for this purpose!)
Quote ELL Sonnhammer, G von Heijne, and A Krogh: A hidden Markov model for predicting transmembrane helices in protein sequences. Proc of the Sixth Intern Conf on Intelligent Systems for Molecular Biology (ISMB98), 175-182, 1998
Authors Anders Krogh (CBS, Copenhagen, Denmark)
Contact Anders Krogh ( )
Server TopPred
Site (URL)
About Prediction of location and orientation of transmembrane helices through an advanced use of hydrophobicity patterns, and by applying the 'positive-inside' rule.
  1. G von Heijne: Membrane Protein Structure Prediction, Hydrophobicity Analysis and the Positive-inside Rule. J. Molecular Biology, 225, 487-494, 1992
  2. M Cserzo, E Wallin, I Simon, G von Heijne, and A Elofsson: Prediction of transmembrane alpha-helices in prokaryotic membrane proteins: the dense alignment surface method. Protein Engineering, 10, 673-676, 1997
Authors Erik Wallin, and Gunnar von Heijne (Stockholm Univ, Sweden)
Contact Erik Wallin (
Server DAS
Site (URL)
About Prediction of location of transmembrane helices through an advanced use of hydrophobicity patterns.
Quote M Cserzo, E Wallin, I Simon, G von Heijne, and A Elofsson: Prediction of transmembrane alpha-helices in procariotic membrane proteins: the Dense Alignment Surface method. Protein Engineering, 10, 673-676, 1997
Authors Miklos Cserzo, Istvan Simon (both Academy of Sciences, Budapest, Hungary), Erik Wallin, Gunnar von Heijne, Arne Elofsson (Stockholm Univ, Sweden)
Contact Miklos Cserzo (

Threading (detecting remote homologues in PDB)

Threading servers search through databases of proteins of known structures (subsets of PDB), and detect similiarities between proteins which are too weak to be inferred from simple sequence alignment techniques. (The 'trick' that allows to intrude below the twilight zone of 'simple' similarity detection is using the information contained in the known structures of the proteins compared to your sequence.)
Note: detection of remote homologues is more of an art than of a solved problem. Most of the results returned are supposedly wrong! Thus, you have to gather independent evidence to be able to believe in a particular result.
Server FRSVR
Site (URL)
About Prediction-based threading, also incorporating purely sequence-based database searches.
  • Include H3P2: requests to include the results of the H3P2 prediction.
  • Include PROFILESEARCH: requests to include the results of the PROFILESEARCH program (Smith-Waterman using a profile of the multiple alignment).
  1. D Fischer, and DA Eisenberg: Fold Recognition Using Sequence-Derived Properties. Protein Science, 5, 947-955, 1996
  2. A Elofsson, D Fischer, DW Rice, S LeGrand, and DA Eisenberg: Study of Combined Structure-sequence Profiles. Folding and Design, 1, 451-461, 1996
Authors Daniel Fischer (Ben Gurion Univ of the Negev, Israel)
Contact Daniel Fischer ( )
Server SAMt98
Site (URL)
About A new hidden Markov model method (SAM-T98) for finding remote homologs of protein sequences is described and evaluated. The method begins with a single target sequence and iteratively builds a hidden Markov model (HMM) from the sequence and homologs found using the HMM for database search. SAM-T98 is also used to construct model libraries automatically from sequences in structural databases.
Quote Kevin Karplus, Christian Barrett, and Richard Hughey:Hidden Markov Models for Detecting Remote Protein HomologiesBioinformatics, 14, 846-856, 1998
Authors Kevin Karplus, Christian Barrett, and Richard Hughey (UCSD, Santa Cruz, USA)
Contact SAM-INFO (

Homology modelling

Homology modelling predicts the three-dimensional structure for your sequence based on the similarity of that sequence to a protein of experimentally determined structure. Since the assumption of the method is that the backbone of your protein and the similar one of known structure (referred to as 'the template') are identical, the sequence similarity between the two has to be significant!
Note: if there is no similar structure in PDB, 3D structure can NOT (repeat NOT) be predicted by ANY method at this moment!
Site (URL)
About SWISS-MODEL is an Automated Protein Modelling Server running at the GlaxoWellcome Experimental Research in Geneva, Switzerland (click here for more details on how the method works).
  1. M C Peitsch: Protein Modelling by E-mail.Bio/Technology, 13, 658-660, 1995.
  2. M C Peitsch: ProMod and Swiss-Model: Internet-based tools for automatedcomparative protein modelling.Biochem Soc Trans, 24, 274-279, 1996.
  3. N Guex, and M C Peitsch:SWISS-MODEL and the Swiss-PdbViewer:An environment for comparative protein modelling.Electrophoresis, 18, 2714-2723, 1997.
Authors Manuel Peitsch, , Torsten Schwede, and Nicolas Guex (Glaxo, Geneva, Switzerland)
Contact Nicolas Guex (
Server CPHmodels
Site (URL)
About CPHmodels is a collection of methods and databases developed to predict protein structures. It currently consists of the following tools: Sowhat: A neural network based method to predict contacts between C-alpha atoms from the amino acid sequence. RedHom: A tool to find a subset with low sequence similarity in a database. Databases: Subsets of the Brookhaven Protein Data Bank (PDB) database with low sequence similarity produced using the RedHom tool.
Quote O Lund, K Frimand, J Gorodkin, H Bohr, J Bohr, J Hansen, and S Brunak:Protein distance constraints predicted by neural networks and probability density functions. Protein Engineering, 10, 1241-1248, 1997
Authors Ole Lund (CBS, Copenhagen, Denmark)
Contact Kristoffer Rapacki (

Formats for submitting sequence or alignment
The following formats are valid for submitting your sequence, or alignment.
Note: we recommend to use either of the formats in bold print!