Method at a Glance

WESA is a meta-predictor for solvent accessibility of residues from protein sequences. It is based on an ensemble of five methods: Bayesian statistics (BS), multiple linear regression (MLR), decision tree (DT), neural network (NN), and support vector machine (SVM). The final prediciton for each residue is based on a weighted sum of the individual predictions. A residue is predicted either as buried or exposed (defined as having a surface area greater than 20% the maxmimum area for that type of amino acid). WESA has an expected accuracy of 80%.

  • Chen, H-L and Zhou, H-X. Prediction of solvent accessibility and sites of deleterious mutations from protein sequence. Nucl. Acids Res. 33, 3193-3199 (2005).
  • Shan, Y., Wang, G., and Zhou, H.-X. Fold recognition and accurate query-template alignment by a combination of PSI-BLAST and threading. Proteins 42, 23-37 (2001).

  Note: The length of sequence can be submitted here is limited to 2000 amino acids.

Enter the sequence of your protein to get predicted solvent accessibility

If you have more than a few sequences and would like to run WESA in batch mode, click here.

  • Type an arbitrary name for referencing your submission:
  • Enter your email address, where the prediction will be sent to:
  • Paste the amino acid sequence (in one-letter code):

  • Home

  • Output format

  • About the figures

  • cons-PPISP


  • Download dataset

  • View Stats