conSSert

conSSert - Consensus Secondary Structure Prediction

conSSert is a SVM-based consenus secondary structure prediction method. Predictions from four popular secondary structure models (PSSpred, PSIPRED, Raptor-X, and SPINE-X) are integrated through the use of SVM models to produce highly accurate predictions, especially with regard to Q2-EH.

conSSert takes a single protein amino acid sequence in FASTA format as input, and returns two output files, via email, once the prediction is complete. A PDF file (conSSert.pdf) containing a color-coded (red is strand and blue is helix) comparison of the input sequence and the three-class secondary structure prediction is provided. Additionally, a plain text document (output.txt) containing a more detailed representation of the data is also provided. In the provided table the column nomenclature is as follows:


Example

Input sequence

>DEFB1_HUMAN Beta-defensin 1
MRTSYLLLFTLCLLLSEMASGGNFLTGLGHRSDHYNCVSSGGQCLYSACPIFTKIQGTCYRGKAKCCK

Output Files

conSSert.pdf
output.txt


Submit Protein Amino Acid Sequence

(optional)
 

Publications

Kieslich, C. A., Smadbeck, J., Khoury, G. A., and Floudas, C. A. conSSert: Consensus SVM model for accurate prediction of ordered secondary structure. J. Chem. Inf. Model., 56 (3), pp 455–461, 2016.