NeBcon (Neural-network and Bayes-classifier based contact prediction) is a hierarchical algorithm for sequence-based protein contact map prediction. It first uses the naive Bayes classifier theorem to calculate the posterior probability of eight machine-learning and co-evoluation based contact prodiction programs (SVMSEQ, BETACON, SVMcon, PSICOV, CCMpred, FreeContact, MetaPSICOV, and STRUCTCH). Final contact maps are then created by neural network machine that trains the posterior probability scores with intrinsic structural features from secondary structure, solvent accessibility, and Shannon entropy of multiple sequence alignments.

NeBcon On-line (view an example of NeBcon output)

Cut and paste your sequence ([20, 1000] residues in FASTA format) below: Example input

Or upload the sequence from your local computer:

Email: (mandatory, where results will be sent to)

ID: (optional, your given name of the protein)

Download package:

The standalone NeBcon package can be downloaded from NeBconpackage.tar.gz. To install the package, follow the instrunctions below:
1. Decompress the NeBconpackage.tar.gz with the command: tar -zxvf NeBconpackage.tar.gz
2. After decompressing, read the README.txt file, which is available in NeBconpackage, for further instructions.

References:

B He, SM Mortuza, Y Wang, H Shen, Y Zhang. NeBcon: Protein contact map prediction using neural network training coupled with naïve Bayes classifiers. Bioinformatics, 33: 2296-2306 (2017). [PDF] [Support Information]

Supporting Materials: A list of the training dataset (517 non-homolgous proteins) and the test dataset (98 proteins) can be found at here.