DOCKovalent cysteine inhibitor design tutorial

From DISI
Revision as of 22:22, 18 July 2018 by Xiaobo wan (talk | contribs)
Jump to navigation Jump to search

This was written on April 4, 2018.

This tutorial is for designing linkers for a covalent inhibitor and is supplement the work in preparation (Wan et al 2018).


These file are in the /mnt/nfs/home/xiaobo/UCSF_scripts/2018-4-3-covlanet_lysine_wiki-tutorial


Step 1. Custom Ligand and Library Generation

1/Custom Ligand / Library Generation

cd 1-Custom-Ligand-Library-Generation 

library 1:  search the acrylamide in the ZINC15 database (ask John to put your builded library in ZINC15)
login into http://zinc15.docking.org/patterns/home/, search acrylamide in pattern
found 1: Acrylamide-Terminal, [CD1]=[CD2]-C(=O)-[NX3]    Purchase is 84576
File:    /nfs/home/xiaobo/UCSF_scripts/2018-7-17-covalent_cys_wiki-tutorial/1-Custom-Ligand-Library-Generation/acrylamide/library1-70587-ZINC15-acrylamide-library.smi
db2 file in  /mnt/nfs/ex9/work/xiaobo/new_covalent_lib/acrylamides/lib1

library 2: aldehyde-based-cyanoacrylamides
Search the aldehyde from ZINC15, and only-single-aldehyde-aromatic-for-sale+bb.smi 145960
one step synthesis
       python aldehyde-to-cyanoacryl.py only-single-aldehyde-aromatic-for-sale+bb.smi
       python step3-remove_doubles.py reaction_nocorina_out.ism
File:
       /nfs/home/xiaobo/UCSF_scripts/2018-7-17-covalent_cys_wiki-tutorial/1-Custom-Ligand-Library-Generation/aldehyde-based-cyanoacrylamides/145956-aldehyde-based-cyanoacrylamides.smi
db2 file  in /mnt/nfs/ex7/work/xiaobo/new_covalent_lib/2017-6-8-cyanoacrylamide

library 3: ~184,900 Enamine acids + Boc-diamine + acrylic acid library
filter3-selected_acids_2150.smi the most common 2150 Enamine acids fro Enamine
83-Boc_diamines.smi  the most common 83 Boc from Enamine
two step synthesis
python step1-SN1-diamines-CO2H.py filter3-selected_acids_2150.smi 83-Boc_diamines.smi
python step2-reaction-acrylic-acid.py in.smi
pyton step3-remove_doubles.py in2.smi
final file: /nfs/home/xiaobo/UCSF_scripts/2018-7-17-covalent_cys_wiki-tutorial/1-Custom-Ligand-Library-Generation/acids-Boc-acrylic-acid/final-acids-Boc-acrylic-acid.smi    184900
db2 file in  /mnt/nfs/ex7/work/xiaobo/2017-6-30-acids-Boc-acrylic-acid/acd1
             /mnt/nfs/ex7/work/xiaobo/2017-6-30-acids-Boc-acrylic-acid/acd2

library 4: ~145,508 Sulfonyl Chloride + Boc-diamine + acrylaic acid library
filter4-1677.sulfonyl_chlorides.smi the most common 1677 Enamine sulfonyl_chlorides fro Enamine
83-Boc_diamines.smi  the most common 83 Boc from Enamine
two step synthesis
python SN1-diamines-CO2H.py filter4-1677.sulfonyl_chlorides.smi 83-Boc_diamines.smi
python step2-reaction-acrylic-acid.py in.smi
pyton step3-remove_doubles.py in2.smi
final file :  /nfs/home/xiaobo/UCSF_scripts/2018-7-17-covalent_cys_wiki-tutorial/1-Custom-Ligand-Library-Generation/sulfonyl_chloride_Boc-acrylic-acid/final-sulfonyl_chloride_Boc-acrylic-acid.smi
db2 file in  /mnt/nfs/ex7/work/xiaobo/2017-6-30-sulfonyl_chloride_Boc-acrylic-acid/suc1
             /mnt/nfs/ex7/work/xiaobo/2017-6-30-sulfonyl_chloride_Boc-acrylic-acid/suc2


The no_double_out.ism was used to generate db2 file for covalent docking 

log into gimel
setenv DOCKBASE /mnt/nfs/home/xiaobo/combine_docknormal_dock_covalent_3.7_and_tart/DOCK_from_githup_2016_5_27
setenv DOCKBASE /mnt/nfs/home/xiaobo/combine_docknormal_dock_covalent_3.7_and_tart/DOCK_from_githup_2016_5_27
/nfs/soft/tools/utils/qsub-slice/qsub-mr-meta -tc 50 --map-instance-script "/nfs/soft/tools/utils/qsub-slice/qsub-mr-map.sh" -s $BUILD_ENVIRONMENT -l 1 no_doubles_out.ism $DOCKBASE/ligand/generate/build_database_ligand.sh --no-db --no-solv --no-mol2 --single --covalent

Step 2 Protein preparation (different lysine rotamers)

2/Protein preparation (different lysine rotamers) 

cd 2-Protein-preparation-different-lysine-rotamers

find the modification lys number in the PDB 

echo "5K9I-B-X44      B       295">>lys.list
bash step0_prepare_build_system.sh  5K9I-B-X44

In the window of chimera, select all of the 27 lysine rotamers and click the button of OK. Reselect all the lysine rotamers in the PDB structure, and the save to PDB format LYS-5K9I-B-X44.pdb Then, to generate all 28 structure folds, and then automatically calculate the steric clash with nearby residues, and select the rotamer with no steric clashes. This script will also calculate the nearest atom of in the compound to the lysine NZ atom 

bash step1_run_build_system.sh 5K9I-B-X44 
results
5K9E-B-X44      SBH     2.038
5K9B-B-X44      SBH     2.321
5K9I-B-X44      OBI     2.949
5K9L-B-X44      SBH     4.683
5K9R-B-X44      OBI     4.925

Each folder contains rec.pdb and xtal-lig.pdb

For each folder 

bash step1_DOCKINV.blastermaster.sh 5K9I-B-X44 box_margin(10) 1(covalent docking)

box_margin is defined from the center of the xtal-lig.pdb file

Step 3 modify the INDOCK parameters for saving multiple poses

cd 3-modify-the-INDOCK-parameters

change the default parameters for covalent docking 

 bump_maximum                  100
 bump_rigid                    100
 number_save                   1000
 number_write                  1000
 molecules_maximum             100000
 bond_len                      1.61
 bond_ang1                     121.02
 bond_ang2                     107.36
 len_range                     0.0
 len_step                      0.1
 ang1_range                    20.0
 ang2_range                    20.0
 ang1_step                     5
 ang2_step                     5
 check_clashes                 no
 per_atom_scores               yes

Step 4 run the covalent docking in gimel

cd 4-run-the-covalent-docking

contain a pharmacophore filter ( exclusion criteria that ligands should form hydrogen bonds with the kinase hinge region, and the shared pyrimidine 3-aminopyrazole scaffold should be within 2 Å compared to the crystal conformation)

Prepare 

1)the modified INDOCK file INDOCK.bump1000000000000.pose1000.20.5.5
2)the gate residue file (define the two residue in the SRC kinase domain MET341  VAL399)

Input file : 

1) the list different structure folders (5K9A-B-X44,5K9A-C-X44)
2) the ligand library folder name (lib1)
3) the linker name list (lib1.list)
 bash /mnt/nfs/home/xiaobo/UCSF_scripts/2018-4-3-covlanet_lysine_cys_wiki-tutorial/4-run-the-covalent-docking/qsub_multipe_jobs structure-list lib1  lib1.list

Step 5 Analysis and combine the top1 pose from different structures

 cd 5-Analysis-and-combine-the-top1-poses-from-different-structures

after the covalent docking, analyze the docking results 

 bash step1_extract_the_best_score.sh  structure-list lib1  lib1.list
      Inputfile :
      1) the list different structure folders (5K9A-B-X44,5K9A-C-X44)
      2) the ligand library folder name (lib1)
      3) the linker name list (lib1.list)

combine the docking results 

       bash step3_combine-best-energy.sh structure-list lib1

       Input file :
       1) the list different structure folders (5K9A-B-X44,5K9A-C-X44)
       2) the ligand library folder name (lib1)
       Output file:
       1)sort.final.combine-new.aura-A-X63.list.dat    rank all of the top1 pose for each linker
       2)submit.new.aura-A-X63.list.dat                the name of structure file and linker

Step 6 Run the minimization and MM/GBSA rescoreing

 cd 6-Run-the-minimization-and-MMGBSA-rescoring

First, check the protonation state of each linker after when using the chimera to add hydrogen second, the different H position of linkers will result in the different labelling number of the attached NH of lysine residue prepare the list for each linker containing two informations in XO44.charge.list file (default:xabs 1 1) 

 bash step7_fix_prolem_resubmit_MMPBSA.minimization.sh INDOCK.bump1000000000000.pose1000.20.5.5-xo4E-A-X44-X44-meta-xaaa-1-mini_end_GB

INDOCK.bump1000000000000.pose1000.20.5.5-xo4E-A-X44-X44-meta-xaaa-1-mini_end_GB is the folder for runing minimization

after minimization, then run the AMBER MMGBSA rescoring bash step10_fix_prolem_resubmit_MMPBSA_score.sh INDOCK.bump1000000000000.pose1000.20.5.5-xo4E-A-X44-X44-meta-xaaa-1-mini_end_GB

extract the scoring number for each linker 

 bash step6_resubmit.extract_GBscore.sh list

the list contains (INDOCK.bump1000000000000.pose1000.20.5.5-xo4E-A-X44-X44-meta-xaaa-1-mini_end_GB)

Step 7 analyze the final pose by chimera

 cd 7-analyze-the-final-pose-by-chimera

first sort the linker according to the MMGBSA score 

 cat MMGBSA.list | sort -nk 2 >sort.MMGBSA.list

 1-extract the pose without the protein
 perl fix-step3_extract_best_score_combinepdb_after_minimize.pl sort.MMGBSA.list
 2-extract the pose with the protein
 perl fix-step4_extract_best_score_combinepdb_after_minimize_with_rec.pl sort.MMGBSA.list

using the chimera to visualize these poses and select the final linker (save to PDB file) 

 save the linker viewdock state: P
 perl step0-filter_by_the_chimera.pl pdb  to extract the final poses
Retrieved from "http://wiki.docking.org/index.php?title=DOCKovalent_cysteine_inhibitor_design_tutorial&oldid=10829"