These scripts are designed to run various parts of candock in a modular fashion. For example: they can be used to generate the fragments for docking without actually doing any docking. These modular can be run independently of each other, and dependencies between modules are taken care of automatically ( IE binding site identification will take place before docking of fragments ). The list of modules is given below:
The modules can be used several different ways and all of these ways are controlled by the same set of variables.
Path containing the candock directories.
Path containing all of the the scripts! It’s recommended that one exports this variable in their .bash_profile file. Otherwise, the script will attempt to determine this automatically, which may not work in all cases. You have been warned.
Version of candock to use.
Number of CPUs to use.
File containing the protein that one wishes to dock to.
File containing all the ligands that one wishes to use.
File to read the binding site from. If this file is not present or is empty, then the binding site will be determined automatically and saved to this file.
File containing all the ligands that have been tagged with which fragments they consist of. If this file does not exist or is empty, then the ligands in $ligands will be tagged as such and saved to this file.
Directory containing the docked fragments. If it does not exist, then the fragments given in $prep will be docked to the binding site given in $bsite.
Flag controlling rigid vs flexible docking. If set to 0, then rigid docking will occur. If set to any value greater than 0, then flexible docking will occur.
Number (as a percentage) of each of the highest scored seeds to use. Default is 0.02 which means 2%.
Maximum number of iterations to preform while linking.
Maximum of clustered confirmations to link.
There are three ways to use the modules. Each way has advantages and disadvantages and the correct
Each module exists as both a Bash script and a Bash function. This method is the quickest and dirtiest way of using the modules. It should not be used for long jobs and is best for quickly checking the modules work properly with a given set of variables. To use, start by using the following to load the functions in to the Shell environment source $MCANDOCK_MOD_PATH/load_variables.sh. Now you can invoke a module by simply typing the module name as you would a program name. For example, one could simply type bsite to do binding site identification. To change a variable, simply say varname=varvalue. To change the receptor, for example, use the following: CANDOCK_receptor=1aaq.pdb.
This method is the recommended method for invoking a module when using a Lab or local machine. Do not use it for jobs that are to be run on the cluster. To invoke a module, simply add $MCANDOCK_MOD_PATH to your $PATH and type the module name. For example, just type prep_fragments.sh and your off and running! To change a variable, you must export it first. Do so like the following example for ligands: export CANDOCK_ligands=new_drug.mol2.
If you’re using the modules on an RCAC cluster, you must use this method if your jobs is to run for more than a few minutes. It is the most convoluted, but the most powerful. To start, jobs must be submitted using qsub and the full path to each module must be given. Variables are given to qsub as comma separated equalities using the -v option. For example: qsub -v var1=value1,var2=value2. Do not use spaces to separate values. Alternately, one can export variables and use qsub -V. If using a different $MCANDOCK_MOD_PATH than the default, ensure that this variable is exported properly as PBS blocks the scripts ability to determine it’s location. For an interactive job, export $MCANDOCK_MOD_PATH before using the script.
So, the above was really confusing ~and badly written~ - so now there’s an easier way! (tm). The submit_candock_module.sh command simplifies things by running qsub for you. To run dock_fragments, for example, use submit_candock_module.sh dock_fragments. You can pass qsub arguments as well, thus submit_candock_module.sh dock_fragments -v var1=value1 -l myarg=whatever is valid. Note that -V is passed automatically, so make sure your environment is setup properly!
The following will create a seeds database on the standby queue with 20 cores.
cd $RCAC_SCRATCH
mkdir flt3
cd flt3
dlrcsb.pl 4xuf > 4xuf.pdb
grep '^ATOM' 4xuf.pdb | grep ' A ' > receptor.pdb
cp /depot/gchopra/data/seeds_data_base/cando-ligands-3d.mol2 ./ligands.mol2
submit_candock_module.sh dock_fragmentsThe following will docking a small number of ligands to a single protein. It will have 200 hours to complete and will run in the gchopra queue.
cd $RCAC_SCRATCH
mkdir working_dir
cd working_dir
cp /path/to/my/protein.pdb ./receptor.pdb
cp /path/to/my/drugs.mol2 ./ligands.mol2
submit_candock_module.sh link_fragmentscd $RCAC_SCRATCH
mkdir working_dir
cd working_dir
mkdir structures compounds seeds_database docking
cd structures
# Do this for all pdb codes you want to dock against
dlrcsb.pl 1PDB > 1pdb.pdb
dlrcsb.pl 2PDB > 2pdb.pdb
# .....
basename -s .pdb -a *.pdb > ../all.lst
# You can also place your own binding sites here, named
# 1pdb.cen, 2pdb.cen, etc
cd ../compounds
cp /path/to/directory/with/your/ligands.mol2 .
# If there's a few ligands ( <100ish )
prep_fragments.sh
# A lot of fagmenets
submit_candock_module.sh prep_fragments
cd ../seeds_database
dock_multiple.sh dock_fragments
# wait for all jobs to finish
cd ../docking
mkdir -p top_0.02/all_conf
cd top_0.02/all_conf
dock_multiple.sh link_fragmentsStarting Input Files:
| Option Name | Default | Description |
|---|---|---|
| receptor | receptor.pdb | Receptor filename |
| ligand | ligands.mol2 | Ligand filename |
Probis (binding site indentification) Options:
| Option Name | Default | Description |
|---|---|---|
| neighb | false | Allow only ligands that are in the similar regions according to REMARKs |
| num_bsites | 3 | Maximum number of predicted (or given) binding sites to consider for docking |
| centroid | None | Filename for reading and writing centroids |
| names | bslibdb/data/names | Directory with ligand names |
| lig_clus_file | ligand_clusters.pdb | Ligand clusters found by ProBiS are outputted to this file |
| nosql | probis.nosql | NoSql-formatted ProBiS alignments output file |
| probis_min_z_score | 2.5 | Minimium z-score of ligands to be considered in clustering |
| probis_clus_rad | 3.0 | Cluster radius for predicted ligands byprobis |
| srf_file | probis.srf | File for storing the protein surface calculated by probis. |
| z_scores_file | z_scores.pdb | Binding site z-scores are outputted to this file |
| probis_min_pts | 10 | The minimum number of points (for predicted ligands) required to form a cluster |
| json | probis.json | Json-formatted ProBiS alignments outputfile |
| bio | bslibdb/data/bio | Directory with ProBiS-ligands bio database |
| bslib | bslibdb | Read binding sites library from this directory |
| centro_clus_rad | 3.0 | Cluster radius for centroid centers |
| jsonwl | probis_with_ligands.json | Json-formatted ProBiS alignments with transposed ligands output file |
Ligand Fragmention Options:
| Option Name | Default | Description |
|---|---|---|
| seeds | seeds.txt | Read unique seeds from this file, if itexists, and append new unique seeds if found |
| max_num_ligands | 10 | Maximum number of ligands to read in one chunk |
| prep | prepared_ligands.pdb | Prepared small molecule(s) are outputted to this filename |
| seeds_pdb | seeds.pdb | File to save full seeds into. |
Fragment Docking Options:
| Option Name | Default | Description |
|---|---|---|
| top_seeds_dir | None | Directory for saving top docked seeds |
| top_seeds_file | top_seeds.pdb | Top seeds output file |
| clus_rad | 2.0 | Cluster radius for docked seeds |
| num_univec | 256 | Number of unit vectors evenly distributed on a sphere for conformation generation |
| gridpdb_hcp | gridpdb_hcp.pdb | Grid pdb hcp file for output |
| clusterfile | clustered_seeds.txt | Clustered representative docked-seed conformations output file |
| excluded | 0.8 | Excluded radius |
| conf_spin | 10 | Spin degrees for conformation generation |
| grid | 0.375 | Grid spacing |
| max_frag_radius | 16.0 | Maximum fragment radius for creating the initial rotamers |
| interatomic | 8.0 | Maximum interatomic distance |
Scoring Function Arguments:
| Option Name | Default | Description |
|---|---|---|
| step | 0.01 | Step for spline generation of non-bonded knowledge-based potential [0.0-1.0] |
| obj_dir | obj | Output directory for objective functionand derivatives |
| comp | reduced | Atom types used in calculating reference state ‘reduced’ or ‘complete’(‘reduced’ includes only those atom types present in the specified receptorand small molecule, whereas ‘complete’ includes all atom types) |
| potential_file | potentials.txt | Output file for potentials and derivatives |
| func | radial | Function for calculating scores ‘radial’ or ‘normalized_frequency’ |
| dist | data/csd_complete_distance_distributions.txt | Select one of the interatomic distance distribution file(s) provided with thisprogram |
| cutoff | 6 | Cutoff length [4-15]. |
| scale | 10.0 | Scale non-bonded forces and energy for knowledge-based potential [0.0-1000.0] |
| ref | mean | Normalization method for the reference state (‘mean’ is averaged over all atomtype pairs, whereas ‘cumulative’ is a summation for atom type pairs) |
Forcefield and Minimization Options:
| Option Name | Default | Description |
|---|---|---|
| gaff_dat | data/gaff.dat | Gaff DAT forcefield input file |
| max_iter | 10 | Maximum iterations for minimization during linking |
| pos_tol | 0.00000000001 | Minimization position tolerance in Angstroms - only for KB |
| gaff_xml | data/gaff.xml | Gaff XML forcefield and ligand topologyoutput file |
| mini_tol | 0.0001 | Minimization tolerance |
| update_freq | 10 | Update non-bond frequency |
| water_xml | data/tip3p.xml | Water XML parameters (and topology) input file |
| fftype | kb | Forcefield to use ‘kb’ (knowledge-based) or ‘phy’ (physics-based) |
| max_iter_final | 100 | Maximum iterations for final minimization |
| amber_xml | data/amber10.xml | Receptor XML parameters (and topology) input file |
Fragment Linking Options:
| Option Name | Default | Description |
|---|---|---|
| docked_clus_rad | 2.0 | Cluster radius between docked ligand conformations |
| spin | 60 | Spin degrees to rotate ligand. Allowed values are 5, 10, 15, 20, 30, 60, 90 |
| max_allow_energy | 0.0 | Maximum allowed energy for seed conformations |
| docked_dir | docked | Docked ligands output directory |
| max_clique_size | 3 | Maximum clique size for initial partialconformations generation |
| cuda | 1 (Implicit) | (=false) Enable cuda iterative linker during linking |
| max_possible_conf | -1 | Maximum number of possible conformations to link (-1 means unlimited) |
| link_iter | 1000 | Maximum iterations for linking procedure |
| upper_tol_seed_dist | 2.0 | Upper tolerance on seed distance for getting initial conformations of dockedfragments |
| lower_tol_seed_dist | 2.0 | Lower tolerance on seed distance for getting initial conformations of dockedfragments |
| max_num_possibles | 200000 | Maximum number of possibles conformations considered for clustering |
| tol_seed_dist | 2.0 | Tolerance on seed distance in-between linking |
| top_percent | 0.05 | Top percent of each docked seed to extend to full molecule |
| iterative | 1 (Implicit) | (=false) Enable iterative minimization during linking |
| clash_coeff | 0.75 | Clash coefficient for determining whether two atoms clash by eq. dist12 s< C * (vdw1 + vdw2) |
Automated Design Options:
| Option Name | Default | Description |
|---|---|---|
| change_terminal_atom | None | Change non-hydrogen atoms that terminate chains to given atoms. Multiple atoms can be given. |
| target_dir | targets (Implicit) | Directory containing PDB files. These are docked against and labeled as targets. |
| antitarget_dir | atargets (Implicit) | Directory containing PDB files. These are docked against and labeled as antitargets |
| force_seed | None | Force addition of a certain seed from seeds.pdb. Multiple seeds can be given |
| add_single_atoms | None | Change hydrogens to given atoms. Multiple atoms can be given. |
| target_linking | true | Should the ligands be linked for target |
| antitarget_linking | true | Shoutd the ligands be linked for antitargets |
| seeds_till_bad | -1 | Number of times a seed must be present in the top_seeds for antitargets until it is removed from the good list |
| fragment_bag | fragment_bag.mol2 (Implicit) | Additional fragments to be added to seeds.pdb |
| seeds_to_avoid | 50 | Number of seeds from seeds.pdb to be considered for removal from determined from seeds_to_add |
| seeds_till_good | -1 | Number of times a seed must be present in the top_seeds for targets until it is considered for addition |
| fragment_mol | fragment_mol.mol2 (Implicit) | Additional fragments to be added to seeds.pdb without rotatable bonds beingcut. |
| seeds_to_add | 50 | Number of seeds from seeds.pdb to be considered for addition to the ligands in prepared_ligands.pdb |