Data for: BetaDL: a Protein Beta-sheet Predictor Utilizing a Deep Learning Model and Independent Set Solution

=======================================
BetaDL: a Protein Beta-sheet Predictor Utilizing a Deep Learning Model and Independent Set Solution
Prediction of beta-sheet structures (beta-residue contacts, beta-strand pairs, beta-strand pairing directions)
=======================================
Knowledge Engineering Research Group (KERG), Department of Computer Engineering,
Faculty of Engineering, Ferdowsi University of Mashhad, Iran.
Home page: https://kerg.um.ac.ir/
Contact: [e-mail]; [e-mail]
=======================================
Input:
In order to run BetaDL the following information must be provided:
1. Protein sequence and secondary structure assignment
2. A residue contact Probabilities matrix
3. Probability model for beta-sheet size respect to the number of beta-strands in proteins

Output:
BetaDL will predict the following information about beta sheet structure:
1. Beta-residue contact maps
2. Beta-strand pairs
3. Beta-strand pairing directions
=======================================
Installation (Windows Version) and usage:
1. Unzip "BetaDL_package.rar" into the "BetaDL_package" folder
2. Compile and build the source codes in "BetaDL" folder and move "BetaDL.exe" to the BetaDL_package
3. Provide input files:
Residue contact Probabilities in the "residue contact probabilities" folder, e.g. "\residue contact probabilities\1e5ka.GCNN"
Sequence and secondary structure in the "secondary structures" folder, e.g. "\secondary structures\1e5ka.out"
4. Run "BetaDL.exe"
5. Output file:
Beta-residue contact probabilities, e.g. "\test\1e5ka_DeepBBContact.out"
Beta-residue pairwise alignments, e.g. "\test\1e5ka_alignmnet.out"
Beta-residue contact maps, "e.g. \test\1e5ka_contactmap.out"
Beta-sheet structure, e.g. "\test\1e5ka_conformation.out"
****The correct answer for the sample protein (1e5ka) can be found in following address: "\correct_beta_sheet_structure\1e5ka_correct.out"
Enjoy it!
=======================================
For more information please contact:
[e-mail]; [e-mail]