Source code is written in C#, and contains a Visual Studio 2010 solution named 'LigandCentricNetworkModels '.
It includes:
Edge.cs /defines edge property/
ligand.cs /defines ligand/
NetworkConstruction.cs /creates similarity and identity networks/
Node.cs /defines node/
PDBInterface.cs /reads input file provided as txt file and prepares for use/
Program.cs /main/
We also include a jar file to run the application from command line. Java 1.8+ is required to run the application.
java -jar ligandcentnetworkv1.jar "yourinputfile.txt"
A new version for jar file is also included. In this version you can choose a ligand similarity function, among (1) CDK's own similarity and (2) Lingo-based smiles-centric similarity function.
java -jar ligandcentnetworkv2.jar "yourinputfile.txt" 2
To use your own data, you should prepare it as the following input form in a text file,
ProteinID LigandID SMILES
ProteinID LigandID SMILES
....
i.e.,
P25910 MES C1COCC[NH+]1CCS(=O)(=O)[O-]
P0C5C1 SFR C[C@H]([C@H]([C@@H]1NC(=C(S1)[C@H]2CCCO2)C(=O)O)C(=O)O)O
P0AD64 17O c1c2n(nc1C3C(=CNC(=CS3)C(=O)O)C=O)CSC2
....
After preaparing your file, "input.txt", you should place it under the folder "...\LigandCentricNetworkModels \bin\Debug\Text".
Finally, go to LigandCentricNetworkModels.sln, and change the following input parameter in "Program.cs" with the name of your input file:
string filenamePDB = "Text/PDB_complete.txt";
Outputs of the code are six different text files each containing a network table, w therefore they can be easily imported into Cytoscape.
1 "ide_unw.txt", unweighted identity
2 "ide_wei.txt", weighted identity
3 "ide_norm_wei.txt", normalized weighted identity
4 "sim_unw.txt", unweighted similarity
5 "sim_wei.txt", weighted similarity
6 "sim_norm_wei.txt", normalized weighted similarity
Note: Source code requires ChemAxon JChem package for .NET, in order to calculate similarity between ligands (http://www.chemaxon.com/download/jchem-suite/#jchemdotnet).
Jar file uses CDK (Chemistry Development Kit) to compute similarity. (https://github.com/cdk)