CADDSuiteParallelization

You can use a combination of LigandFileSplitter and DockResultMerger to parallelize time-consuming jobs, e.g. docking or rescoring of very many compounds.

An example for the parallelization of IMGDock by factor 15 is shown below.

The factor of parallelization can be chosen as desired in the LigandFileSplitter. By default the output of the LigandFileSplitter extends the input file name by ascending numbers, but it is also possible to specify the filenames of the split file.

DockResultMerger merges the output of all docking jobs and sorts the molecules ascendingly according to a numerical property tag, e.g. the score assigned by docking, contained in the input files.

The above shown parallelization can be run on the command line with something like:

BALL/build/bin/TOOLS/LigandFileSplitter -i valid_compounds.sdf -no 15

will produce 15 files with the file names valid_compounds_0.sdf, valid_compounds_1.sdf, ..., valid_compounds_14.sdf.

BALL/build/bin/TOOLS/LigandFileSplitter -i valid_compounds.sdf -no 15 -o batch1.sdf batch2.sdf batch3.sdf batch4.sdf batch5.sdf batch6.sdf batch7.sdf batch8.sdf batch9.sdf batch10.sdf batch11.sdf batch12.sdf batch13.sdf batch14.sdf batch15.sdf

will produce 15 files with the specified names batch1.sdf to batch15.sdf

... then dock each batch with IMGDock ...

BALL/build/bin/TOOLS/LigandFileSplitter/DockResultMerger -s score -o merged_results.sdf -i dock_result1.sdf dock_result2.sdf dock_result3.sdf dock_result4.sdf dock_result5.sdf dock_result6.sdf dock_result7.sdf dock_result8.sdf dock_result9.sdf dock_result10.sdf dock_result11.sdf dock_result12.sdf dock_result13.sdf dock_result14.sdf dock_result15.sdf