Secondary structure prediction methods attempts to infer the likely secondary structure for a protein based on its amino acid composition and similarity to sequences with known secondary structure. The most recent version of the method, JPred4, employs a series of neural networks trained to predict different secondary structure types from a sequence profile, and when necessary, employs a jury network to identify the most likely secondary structure prediction.

• Drozdetskiy A, Cole C, Procter J & Barton GJ. (2015)
  JPred4: a protein secondary structure prediction server
  Nucleic Acids Research, Web Server issue (first published 15th April 2015)
  http://dx.doi.org/10.1093/nar/gkv332
• Cole C., Barber J.D. and Barton G.J. (2008) The Jpred 3 secondary structure prediction server Nucleic Acids Research 36 W197-W201
• Cuff J. A and Barton G.J (1999) Application of enhanced multiple sequence alignment profiles to improve protein secondary structure prediction Proteins 40 502-511

• If nothing is selected, and the displayed sequences appear to be aligned, then a JPred prediction will be run for the first sequence in the alignment, using the current alignment. Otherwise the first sequence will be submitted for prediction.
• If just one sequence (or a region on one sequence) has been selected, it will be submitted to the automatic JPred prediction server for homolog detection and prediction.
• If a set of sequences are selected, and they appear to be aligned, then the alignment will be used for a JPred prediction on the first sequence selected in the set (that is, the one nearest the top of the alignment window).

Note: JPred secondary structure prediction is a 'non-column-separable' service - predictions are based on the sequence profile of contiguous stretches of amino-acid sequence. A prediction will only be made on the visible parts of a sequence (see hiding columns) as if it were a contiguous polypeptide chain. Prediction accuracy at the hidden column boundaries may therefore be less than indicated by JPred's own reliability score (see below).

The result of a JPred prediction for a sequence is a new annotated alignment window:

The sequence for which the prediction was made is the first one in the alignment. If a sequence based prediction was made then the remaining sequences in the alignment are the aligned parts of homologs which were used to construct a sequence profile for the prediction. If the prediction was made using a multiple alignment, then the original multiple alignment will be returned, annotated with the prediction.

• Lupas_21, Lupas_14, Lupas_28
  Coiled-coil predictions for the sequence. These are binary predictions for each location.
• Jnet Burial
  Prediction of Solvent Accessibility. levels are
  • 0 - Exposed
  • 3 - 25% or more S.A. accessible
  • 6 - 5% or more S.A. accessible
  • 9 - Buried (<5% exposed)
• JNetPRED
  The consensus prediction - helices are marked as red tubes, and sheets as dark green arrows.
• JNetCONF
  The confidence estimate for the prediction. High values mean high confidence. prediction - helices are marked as red tubes, and sheets as dark green arrows.
• JNetALIGN
  Alignment based prediction - helices are marked as red tubes, and sheets as dark green arrows.
• JNetHMM
  HMM profile based prediction - helices are marked as red tubes, and sheets as dark green arrows.
• JNETPSSM
  PSSM based prediction - helices are marked as red tubes, and sheets as dark green arrows.
• JNETJURY
  A '*' in this annotation indicates that the JNETJURY was invoked to rationalise significantly different primary predictions.