dock_template

Here will store my scripts about the dock on x254 to remind myself of some errors

PATH and python env

export PATH="/home/soft/ucsfdock/DOCK-3.7-bbe1a30c/bin:$PATH"
export PATH="/home/soft/ucsfdock/DOCK-3.7-bbe1a30c/docking/DOCK/bin/:$PATH"
export DOCKBASE="/home/soft/ucsfdock/DOCK-3.7-bbe1a30c/"
python=/home/yjcheng/miniconda3/envs/dock37/bin/python

dock preparation and sub jobs

There are two thing to do before dock, INDOCK file and db2s file.

generate INDOCK

cp the protein file, named as rec.pdb and the ligand file, named as xtal-lig.pdb in the same dir, then run this scripts.

export DOCKBASE='/home/soft/ucsfdock/DOCK-3.7-bbe1a30c/'
conda activate dock37
/home/yjcheng/miniconda3/envs/dock37/bin/python $DOCKBASE/proteins/blastermaster/blastermaster.py --addhOptions=" -HIS -FLIPs " -v  

sel ligand database to dock

The main databases we use are zinc15(about 0.6 billion molecules) and NIBS database(about 400 thousand).

construct ZINC15 dock dirs

use constrcut_zinc_dock.sh in 1_database_sel/zinc_template to constrcut all dir need by zinc and sub dock jobs at the same time.

construct NIBS dock dirs

use construct.sh in 1_database_sel/nibs_template to construct and sub all jobs of dock NIBS dataset

PLOP rescores

After dock, we will use PLOP to calculate MM-GBSA to rescore and rerank the top 1% result.

extract top 1% poses

First, we need to extract top 1% result, which can be finished by 2_PLOP_rescore/extract_pose.sh

After this job successfully finished, it is better to remove all *gz.db file in dock dirs and tar all dock dir in one tar file, which save a lot of files.

PLOP rescore

PLOP rescore will calculate the interaction energy of receptor and ligand by MM-GBSA at the snapshot of docked pose. To do this job, we need to finish four steps. We often create a new dir which has poses.mol2 and rec.pdb for next work

split ligands in poses.mol2

In this step, we will split poses.mol2 into single molecule files. For each molecule, we will generate two files. One is a pdb file whose suffix is .het and a mol2 file. They are necessary for next step. We will use 2_PLOP_rescore/mols_splits.qsub to finish it.

generate all ligands parameters

Then, after all jobs in step one finished, for the MM-GBSA calculation, we need the molecule parameters. We use 2_PLOP_rescore/mols_para_gen.qsub too generate them.

MM-GBSA calculation

After all parameters are generated, we will begin MM-GBSA calculation. We use 2_PLOP_rescore/plop_rescore.qsub to construct all complex structure and sub all jobs.

get new top poses

Then, we use 2_PLOP_rescore/get_top_poses.qsub to get new top results.

Post propress of rescore

After rescoring, we will get about 5000 molecules, which seems to bind the target. Now, we can cluster them based on fingerprints, use LUNA to prioritize the interaction and deprioritize based on ligand torsion.

Then use Chimera for visual inspection. ViewDock Tutorial