Download spack program, create a environment for ConveyorLC
git clone -c feature.manyFiles=true https://github.com/spack/spack.git
. spack/share/spack/setup-env.sh
spack env create conveyorlc
spack env activate conveyorlc
Optional, you can add external compilers if you are on the supercomputers.
module load gcc/12.1.1
moduel load mavmpich2
spack compiler find --scope site
spack external find --scope site mvapich2
Setting up for spack.
spack mirror add develop https://binaries.spack.io/releases/develop
spack buildcache keys --install --trust
specify the dependence for the conveyorlc
git clone https://github.com/XiaohuaZhangLLNL/spack4atom spack4atom
spack repo add spack4atom
spack repo list
spack add boost@1.72.0+mpi+program_options+system+filesystem+regex+serialization+thread
spack add hdf5+cxx
spack add conveyorlc@master
build the all porgrams. It might take a while.
spack install
Some useful commands for munally fixing the configuration if any errors occurs in the installation
spack cd -e conveyorlc
spack concretize –force
The following is the environment setting for running ConveyorLC
. <install_directory>/spack/share/spack/setup-env.sh
spack env activate conveyorlc
export LBindData="$(dirname `which CDT3Docking`)/../data/"
export PATH=<Amber_installation_directory>/bin/:$PATH
ConveyorLC depends on two external libraries to complete the compilation: MPI (https://www.open-mpi.org or https://www.mpich.org) and boost (https://www.boost.org).
Please install the MPI first. Users can use open-mpi or mpich. Either one is OK. Then install the boost. You need to enable the boost-mpi build. In the project-config.jam add one line to the end of the file:
using mpi : <path_to_your_MPI_installation_directory>/bin/mpicxx ;
to install Boost library please follow the step in the Boost document.
./bootstrap.sh --prefix=path/to/installation/prefix
./b2 install
On the LC machines, the MPI and Boost are installed. Users can source the environment and then compile the code.
On quartz:
module load boost/1.62.0
The code can be download from:
https://github.com/XiaohuaZhangLLNL/conveyorlc
by git:
git clone [email protected]:XiaohuaZhangLLNL/conveyorlc.git
Installing it by using cmake is straight forward:
cd conveyorlc
cmake . -DCMAKE_INSTALL_PREFIX:PATH=/usr/gapps/bbs/conveyorlc
make
make install
Use git clone to download spack4atom
git clone ssh://git@cz-bitbucket.llnl.gov:7999/xzr/spack4atom.git
Add the spack4atom repo to spack
spack repo add <path to spack4atom repo root>
Double check if the repo has been added correctly.
[zhang30@quartz2306 spack4atom]$ spack repo list
==> 2 package repositories.
atom /g/g92/zhang30/H/test/spack4atom
builtin /usr/gapps/bbs/TOSS-3/spack/var/spack/repos/builtin
Install the ConveyorLC by spack
spack install conveyorlc
To clean up the installation and re-install
spack uninstall conveyorlc
spack clean --all conveyorlc
spack install conveyorlc
Conduit is an open source project from Lawrence Livermore National Laboratory that provides an intuitive model for describing hierarchical scientific data. Conduit has I/O interfacces to HDF5. The purpose to use Conduit in the program is to reduce the number of files.
To install Conduit:
git clone --recursive https://github.com/llnl/conduit.git
cd conduit
mkdir build
cd build
cmake ../src/ -DCMAKE_INSTALL_PREFIX:PATH=<conduit_installation_dir> -DHDF5_DIR=<Path_to_HDF5_library>
Install pipeline with Conduit and HDF5 support:
cmake ../ -DCMAKE_INSTALL_PREFIX=<conveyorlc_installation_dir> -DBOOST_ROOT=<Path_to_Boost_library> \
-DHDF5_ROOT=<path_to_HDF5_library> -DCONDUIT_DIR=<path_to_Conduit_library>
The executables with Conduit and HDF5 support are
CDT1Receptor CDT2Ligand CDT3Docking CDT4mmgbsa
export LBindData=/usr/gapps/bbs/TOSS-3/conveyorlc_10/data
export PATH=/usr/gapps/bbs/TOSS-3/conveyorlc_10/bin:/usr/gapps/bbs/TOSS-3/openbabel.3.1.1/bin:$PATH
export AMBERHOME=/usr/gapps/bbs/TOSS-3/amber18/
export PATH=$AMBERHOME/bin/:$PATH
If use Spack with ConveyorLC, please swapp the environment setting with section 1.2
#MSUB -A bbs
#MSUB -l nodes=1:ppn=16
#MSUB -l walltime=16:00:00
#MSUB -l partition=syrah
#MSUB -m be
export LBindData=/usr/gapps/bbs/TOSS-3/conveyorlc_10/data
export PATH=/usr/gapps/bbs/TOSS-3/conveyorlc_10/bin:/usr/gapps/bbs/TOSS-3/openbabel.3.1.1/bin:$PATH
export AMBERHOME=/usr/gapps/bbs/TOSS-3/amber18/
export PATH=$AMBERHOME/bin/:$PATH
srun -N 1 -n 16 CDT1Receptor --input pdb.list --output out
The program will detect the cavities on the protein surface. It will need reference point to help identify which cavity is the active site. In the program, there are multiple ways to generate such reference point
- Use a ligand in the active site. The ligand doesn’t need to be big to cover the whole space.
- Use a few key residues in the active site to help identify the active site.
- Use user defined box information.
These input information can be specify in the pdb.list. Please use anyone of the three options in the following pdb.list:
# Comments start with '#'
# The first column of the input always is the path/pdb_file
# The second column and third one are optional
# To specify non-standard residues in receptor start with "NonRes:"
# nonstandard residues also separated by '|'
#############################################
# 1. Use a ligand in the active site
# srun -N $nnodes -n $procs CDT1Receptor --input pdb.list --output out --sitebylig on
# in your submitting script use “--sitebylig on”
pdb/protease.pdb SubRes:pdb/ligand.pdb
#############################################
# 2. Use a few key residues
# To specify key residues in the active site to help identify the binding site start with "KeyRes:"
# residues separated by '|' and are in <three-letter-residue-name>.<residue-id>.[<chain-id>] format
# in your submitting scriptt use “--sitebylig on”
# srun -N $nnodes -n $procs CDT1Receptor --input pdb.list --output out --sitebylig on
pdb/corona.pdb KeyRes:TYR.489
#############################################
# 3. Use user defined box information
#-35.5312,24.1875,4.5 is the coordinates of box center
#16,18,16 is the box dimension.
# in your submitting script don’t use “--sitebylig on”
# srun -N $nnodes -n $procs CDT1Receptor --input pdb.list --output out
pdb/corona.pdb DockBX:-35.5312,24.1875,4.5|16,18,16
Here is the description of general format in pdb.list
# 1. Comments start with '#'
#
# 2. The first column of the input always is the path/pdb_file
# The second column and third one are optional
#
pdb/2y2vA_A.pdb
#
# 3. To specify key residues in the active site to help identify the binding site start with "KeyRes:"
# residues separated by '|' and are in <three-letter-residue-name>.<residue-id>[.<chain-id>] format
#
pdb/2y2vA_A.pdb KeyRes:SGB.203|GLU.202.A
#
# 4. To specify non-standard residues in receptor start with "NonRes:"
# nonstandard residues also separated by '|'
# by default it takes off lib files for example SGB (or SBG.O) for SGB.off
# If mol2 file is used, Please specify it. For example SGB.M for SBG.mol2
#
pdb/2y2vA_A.pdb NonRes:SGB
pdb/2y2vA_A.pdb KeyRes:SGB.203|GLU.202.A NonRes:SGB
pdb/2y2vA_A.pdb KeyRes:SGB.203|GLU.202.A NonRes:SGB.O
pdb/2y2vA_A.pdb KeyRes:SGB.203|GLU.202.A NonRes:SGB.O|NAD.M|GTP.M|FMN.O
#
# 5. To specify substrate ligand for site identification start with "SubRes:" and follow by the relative
# path to ligand file
#
pdb/sarinXtalnAChE.pdb SubRes:pdb/ligand.pdb NonRes:SGB
pdb/1a50_protein.pdb SubRes:pdb/1a50_ligand.mol2
#
If you have known the docking site and has ligand in active site you can run CDT1Receptor this way: (for ligand, it can take mol2, pdb, sdf format)
srun -N 4 -n 64 CDT1Receptor --input pdb.list --output out --sitebylig on
in pdb.list:
pdb/5os3_protein.pdb SubRes:pdb/5os3_ligand.mol2
pdb/5os0_protein.pdb SubRes:pdb/5os0_ligand.mol2
pdb/5os2_protein.pdb SubRes:pdb/5os2_ligand.mol2
You also can use "--minimize off " if you don't do the MM/GBSA rescoring
srun -N 4 -n 64 CDT1Receptor --input pdb.list --output out --minimize off --sitebylig on
Sometimes docking box is not accurate or not you want. For example, the Aurora kinase binding site is a wide groove but we may want to docking ligand just part of binding site. There are two ways to change the docking box.
Use ‘--boxExtend’ to extend the box
srun -N 1 -n 4 -ppdebug CDT1Receptor --input pdb.list. --boxExtend 5
by default, program adds 2 angstroms to six directions of bounding box of cavity. The command above adds 5 angstroms (it has effect on all receptors in pdb.list)
You can define the docking box for individual protein in pdb.list by using “DockBx” keyword In pdb.list
pdb/2y2vA_A.pdb KeyRes:SGB.203|GLU.202.A NonRes:SGB
pdb/sarinXtalnAChE_A.pdb KeyRes:SGB.203 NonRes:SGB DockBX:1.034,-2.453,-3.123|22,25,36
“DockBX:1.034,-2.453,-3.123|22,25,36” 1.034,-2.453,-3.123 is docking box centroid coordinates separated by comma 22,25,36 is docking box dimension separated by comma
If DockBX is defined for a receptor, the grid calculation will be skipped for that receptor and use the user defined docking box.
#MSUB -A bbs
#MSUB -l nodes=1:ppn=16
#MSUB -l walltime=16:00:00
#MSUB -l partition=syrah
#MSUB -m be
export LBindData=/usr/gapps/bbs/TOSS-3/conveyorlc_10/data
export PATH=/usr/gapps/bbs/TOSS-3/conveyorlc_10/bin:/usr/gapps/bbs/TOSS-3/openbabel.3.1.1/bin:$PATH
export AMBERHOME=/usr/gapps/bbs/TOSS-3/amber18/
export PATH=$AMBERHOME/bin/:$PATH
srun -N 1 -n 16 CDT2Ligand --sdf pur2.sdf
Requirement for the SDF file:
1. Convert you ligand coordinates into SDF file
2. Add total charge fields for each ligand in SDF file, for example:
> <i_user_TOTAL_CHARGE>
-2
3. After calculation, you can find the ligand parameters in:
scratch/lig/1/LIG.lib, LIG.prmtop, LIG.inpcrd
scratch/lig/2/LIG.lib, LIG.prmtop, LIG.inpcrd
...
1, 2, .. correspond to the sequence of ligands appearing in the SDF file
#MSUB -A bbs
#MSUB -l nodes=4:ppn=36
#MSUB -l walltime=16:00:00
#MSUB -l partition=syrah
#MSUB -m be
export LBindData=/usr/gapps/bbs/TOSS-3/conveyorlc_10/data
export PATH=/usr/gapps/bbs/TOSS-3/conveyorlc_10/bin:/usr/gapps/bbs/TOSS-3/openbabel.3.1.1/bin:$PATH
export AMBERHOME=/usr/gapps/bbs/TOSS-3/amber18/
export PATH=$AMBERHOME/bin/:$PATH
srun -N 4 -n 4 -c 36 CDT3Docking --exhaustiveness 36 --num_modes 10
#MSUB -A bbs
#MSUB -l nodes=1:ppn=16
#MSUB -l walltime=16:00:00
#MSUB -l partition=syrah
#MSUB -m be
export LBindData=/usr/gapps/bbs/TOSS-3/conveyorlc_10/data
export PATH=/usr/gapps/bbs/TOSS-3/conveyorlc_10/bin:/usr/gapps/bbs/TOSS-3/openbabel.3.1.1/bin:$PATH
export AMBERHOME=/usr/gapps/bbs/TOSS-3/amber18/
export PATH=$AMBERHOME/bin/:$PATH
srun -N 1 -n 16 CDT4mmgbsa
Both docking and rescoring support LOCALDIR on quartz. The only thing you need to do is add the following environment variable to you submitting script:
export LOCALDIR=/dev/shm/<Your_OUN>/
The output data structure is different from XML version. The calculated data are all in the scratch
scratch/
rec/
receptor.hdf5
lig/
ligand.hdf5
dock/
dockHDF5/
dock_proc1.hdf5
dock_proc2.hdf5
dock_proc3.hdf5
...
dock_proc<N-1>.hdf5
gbsa/
gbsaHDF5/
gbsa_proc1.hdf5
gbsa_proc2.hdf5
gbsa_proc3.hdf5
...
gbsa_proc<N-1>.hdf5
rec, lig, dock, gbsa are the sub-directories for temporarily storing the individual calculation. All the files will be deleted after individual calculation is completed.
All the necessary data/files for different stage calculations are stored in HDF5 such as receptor.hdf5, ligand.hdf5, dock_procN.hdf5, and gbsa_procN.hdf5 Receptor preparation will only produce one HDF5 - receptor.hdf5 and so does ligand preparation - ligand.hdf5. dock and MM/GBSA will have maximum of N-1 number of the HDF files, where N is the number of MPI tasks used in the calculation. Note that normally number of MPI tasks used in MM/GBSA is much larger than that in the docking calculation.
The HDF5 files can be opened by HDF viewer downloaded from: https://www.hdfgroup.org/downloads/hdfview/
To visualize the HDF5 file:
hdfview <your_hdf5_file>
Note: this is only a viewer. You cannot modify the files. Please use the following python scripts to do that.
The conveyorLC has been installed:
/usr/gapps/aha/quartz/conveyorlc_10/
The python scripts are in:
/usr/gapps/aha/quartz/conveyorlc_10/scripts
The python command to manipulate the HDF5 files:
export spack=/usr/gapps/bbs/TOSS-3/spack/bin/
export cdtpy=/usr/gapps/aha/quartz/conveyorlc_10/scripts/
CDT1Receptor.py has following options
$spack/cdtPython.sh $cdtpy/CDT1Receptor.py -h
python /usr/gapps/aha/quartz/conveyorlc_10/scripts//CDT1Receptor.py -h
usage: CDT1Receptor.py [-h] [-i INFILE] [-o OUTFILE] [-d] [-n RECNAME]
[-dn DELNAME] [-sn SAVENAME] [-c CHECKDATA CHECKDATA]
[-m META]
optional arguments:
-h, --help show this help message and exit
-i INFILE, --in INFILE
receptor HDF5 input file
-o OUTFILE, --out OUTFILE
receptor HDF5 output file
-d, --del delete all paths for failed calculations
-n RECNAME, --name RECNAME
extract meta data and files by receptor name
-dn DELNAME, --deletename DELNAME
delete path by receptor name
-sn SAVENAME, --savename SAVENAME
save data to HDF5 output file by receptor name
-c CHECKDATA CHECKDATA, --checkdata CHECKDATA CHECKDATA
update meta data by protein name and checkpoint file
name (e.g. sarinXtalnAChE checkpoint.txt)
-m META, --meta META extract meta data from HDF5 file to CSV file
CDT2Ligand.py has following options
$spack/cdtPython.sh $cdtpy/CDT2Ligand.py -h
python /usr/gapps/aha/quartz/conveyorlc_10/scripts//CDT2Ligand.py -h
usage: CDT2Ligand.py [-h] [-i INFILE] [-o OUTFILE] [-d] [-l LIGID]
[-n LIGNAME] [-c CLIST CLIST]
optional arguments:
-h, --help show this help message and exit
-i INFILE, --in INFILE
Ligand HDF5 input file
-o OUTFILE, --out OUTFILE
Ligand HDF5 output file
-d, --del delete all paths for failed calculations
-l LIGID, --ligid LIGID
extract data and files by ligand ID
-n LIGNAME, --name LIGNAME
extract data and files by ligand name
-c CLIST CLIST, --clist CLIST CLIST
output ligand name and id map into a file and choose
only successful one or not (e.g. idnameList.txt T[A])
CDT3Docking.py has following options
$spack/cdtPython.sh $cdtpy/CDT3Docking.py -h
python /usr/gapps/aha/quartz/conveyorlc_10/scripts//CDT3Docking.py -h
usage: CDT3Docking.py [-h] [-i INDIR] [-o OUTDIR] [-r RECNAME] [-l LIGID]
[-p PERCENT] [-s SCOREONLY]
optional arguments:
-h, --help show this help message and exit
-i INDIR, --in INDIR receptor HDF5 input file
-o OUTDIR, --out OUTDIR
receptor HDF5 output file
-r RECNAME, --recname RECNAME
receptor name for extracting data and files (also need
ligand id)
-l LIGID, --ligid LIGID
ligand id for extracting data and files (also need
receptor name)
-p PERCENT, --percent PERCENT
select top percent ligands into the HDF5 output file
-s SCOREONLY, --scoreonly SCOREONLY
extract score-only data from HDF5 file to CSV file
CDT4mmgbsa.py has following options
$spack/cdtPython.sh $cdtpy/CDT4mmgbsa.py -h
python /usr/gapps/aha/quartz/conveyorlc_10/scripts//CDT4mmgbsa.py -h
usage: CDT4mmgbsa.py [-h] [-i INDIR] [-o OUTDIR] [-r RECNAME] [-l LIGID]
[-m META] [-d]
optional arguments:
-h, --help show this help message and exit
-i INDIR, --in INDIR receptor HDF5 input file
-o OUTDIR, --out OUTDIR
receptor HDF5 output file
-r RECNAME, --recname RECNAME
receptor name for extracting data and files (also need
ligand id)
-l LIGID, --ligid LIGID
ligand id for extracting data and files (also need
receptor name)
-m META, --meta META extract meta data from HDF5 file to CSV file
-d, --del delete all paths for failed calculations
A tool named hdfx is installed in /usr/gapps/bbs/TOSS-3/hdf_utils/hdfx which can extract information from HDF5 files much faster than python script: You can use hdfx -h to see the options or ask Dan for detailed information
- Get scores
/usr/gapps/bbs/TOSS-3/hdf_utils/hdfx -m gbsaHDF5 > meta_data.txt
- Parse the scores
- Get complex files and ligand files
/usr/gapps/bbs/TOSS-3/hdf_utils/hdfx -k gbsa/lv14/286774/p6 -w temp gbsaHDF5
/usr/gapps/bbs/TOSS-3/hdf_utils/hdfx -k lig/286774 -w temp ligand.hdf5
132542 is lig id, compound id can be retrieved from the meta_data
There will be executables in the ${conveyorlc_install_dir}/bin
PPL1Receptor PPL2Ligand PPL3Docking PPL4mmgbsa PPL4PostProcess PPL4parseXML
For the help information of executables
PPL1Receptor -h
#!/bin/bash
#msub -A bmc
#msub -l nodes=2:ppn=16
#msub -l walltime=16:00:00
#msub -l partition=cab
#msub -m be
export conveyorlc=/usr/gapps/kras/quartz/pipeline/conveyorlc
export LBindData=$conveyorlc/data
export PATH=$conveyorlc/bin:/usr/gapps/kras/quartz/pipeline/bin:$PATH
export AMBERHOME=/usr/gapps/kras/quartz/amber16
export PATH=$AMBERHOME/bin/:$PATH
test -f /usr/gapps/kras/quartz/amber16/amber.sh && source /usr/gapps/kras/quartz/amber16/amber.sh
srun -N 2 -n 24 PPL1Receptor --input pdb.list --output out
#!/bin/bash
#msub -A bmc
#msub -l nodes=10:ppn=16
#msub -l walltime=16:00:00
#msub -l partition=cab
#msub -m be
export LBindData=/usr/gapps/bbs/TOSS-3/conveyorlc_10/data
export PATH=/usr/gapps/bbs/TOSS-3/conveyorlc_10/bin:/usr/gapps/bbs/TOSS-3/openbabel.3.1.1/bin:$PATH
export AMBERHOME=/usr/gapps/bbs/TOSS-3/amber18/
export PATH=$AMBERHOME/bin/:$PATH
test -f /usr/gapps/kras/quartz/amber16/amber.sh && source /usr/gapps/kras/quartz/amber16/amber.sh
srun -N 10 -n 160 PPL1Receptor --sdf ligand.sdf
Requirement for the SDF file:
1. Convert you ligand coordinates into SDF file
2. Add total charge fields for each ligand in SDF file, for example:
> <i_user_TOTAL_CHARGE>
-2
3. After calculation, you can find the ligand parameters in:
scratch/lig/1/LIG.lib, LIG.prmtop, LIG.inpcrd
scratch/lig/2/LIG.lib, LIG.prmtop, LIG.inpcrd
...
1, 2, .. correspond to the sequence of ligands appearing in the SDF file
#!/bin/bash
#msub -A bmc
#msub -l nodes=64:ppn=16
#msub -l walltime=16:00:00
#msub -l partition=cab
#msub -m be
export LBindData=/usr/gapps/bbs/TOSS-3/conveyorlc_10/data
export PATH=/usr/gapps/bbs/TOSS-3/conveyorlc_10/bin:/usr/gapps/bbs/TOSS-3/openbabel.3.1.1/bin:$PATH
export AMBERHOME=/usr/gapps/bbs/TOSS-3/amber18/
export PATH=$AMBERHOME/bin/:$PATH
test -f /usr/gapps/kras/quartz/amber16/amber.sh && source /usr/gapps/kras/quartz/amber16/amber.sh
srun -N 64 -n 64 -c 16 PPL3Docking --recXML PPL1Track.xml --ligXML PPL2Track.xml --exhaustiveness 32
#!/bin/bash
#msub -A bmc
#msub -l nodes=32:ppn=16
#msub -l walltime=16:00:00
#msub -l partition=cab
#msub -m be
export LBindData=/usr/gapps/bbs/TOSS-3/conveyorlc_10/data
export PATH=/usr/gapps/bbs/TOSS-3/conveyorlc_10/bin:/usr/gapps/bbs/TOSS-3/openbabel.3.1.1/bin:$PATH
export AMBERHOME=/usr/gapps/bbs/TOSS-3/amber18/
export PATH=$AMBERHOME/bin/:$PATH
test -f /usr/gapps/kras/quartz/amber16/amber.sh && source /usr/gapps/kras/quartz/amber16/amber.sh
srun -N 32 -n 512 PPL4mmgbsa --comXML PPL3Track.xml
The pipeline will generate a lot of files and can cause problem for the file system. PPL4PostProcess tar-zip all mmgbsa poses for the same ligand into one file.
#!/bin/bash
#msub -A bmc
#msub -l nodes=32:ppn=16
#msub -l walltime=16:00:00
#msub -l partition=cab
#msub -m be
export LBindData=/usr/gapps/bbs/TOSS-3/conveyorlc_10/data
export PATH=/usr/gapps/bbs/TOSS-3/conveyorlc_10/bin:/usr/gapps/bbs/TOSS-3/openbabel.3.1.1/bin:$PATH
export AMBERHOME=/usr/gapps/bbs/TOSS-3/amber18/
export PATH=$AMBERHOME/bin/:$PATH
srun -N 4 -n 64 PPL4PostProcess --comXML PPL3Track.xml
Summray of outputs are in the XML format.
The protein target preparation output.
<?xml version="1.0" ?>
<Receptors>
<!-- Tracking calculation error using XML file -->
<Receptor>
<RecID>1ald</RecID>
<PDBPath>pdb/1ald.pdb</PDBPath>
<PDBQTPath>scratch/com/1ald/rec/1ald.pdbqt</PDBQTPath>
<GBEN>-1.0395E+04</GBEN>
<Site>
<Cluster>3</Cluster>
<Volume>154</Volume>
<Centroid>
<X>-48.2348</X>
<Y>50.1281</Y>
<Z>47.2788</Z>
</Centroid>
<Dimension>
<X>23.1982</X>
<Y>21.3827</Y>
<Z>19.4385</Z>
</Dimension>
</Site>
<Mesg>Finished!</Mesg>
</Receptor>
...
</Receptors>
The ligand preparation output.
<?xml version="1.0" ?>
<Ligands>
<!-- Tracking calculation error using XML file -->
<Ligand>
<LigID>3</LigID>
<LigName>Sorafenib</LigName>
<PDBPath>scratch/lig/3/LIG_min.pdb</PDBPath>
<PDBQTPath>scratch/lig/3/LIG_min.pdbqt</PDBQTPath>
<GBEN>-100.76</GBEN>
<Mesg>Finished!</Mesg>
</Ligand>
...
</Ligands>
The docking calculation output
<?xml version="1.0" ?>
<Complexes>
<!-- Tracking calculation error using XML file -->
<Complex>
<RecID>2y2vA_A</RecID>
<NonStdAAList>
<NonStdAA id="1">SGB</NonStdAA>
</NonStdAAList>
<LigID>2</LigID>
<LogPath>scratch/com/2y2vA_A/dock/2/scores.log</LogPath>
<PosePath>scratch/com/2y2vA_A/dock/2/poses.pdbqt</PosePath>
<Scores>
<Pose id="1">-9.1</Pose>
<Pose id="2">-8.7</Pose>
<Pose id="3">-8.4</Pose>
<Pose id="4">-8.2</Pose>
<Pose id="5">-8.1</Pose>
</Scores>
<Mesg>Finished!</Mesg>
</Complex>
...
</Complexes>
The MM/GSBA calcualtion output
<?xml version="1.0" ?>
<Complexes>
<!-- Tracking calculation error using XML file -->
<Complex>
<RecID>2y2vA_A</RecID>
<LigID>3</LigID>
<PoseID>2</PoseID>
<Vina>-8.7</Vina>
<GBSA>29.9987</GBSA>
<Mesg>Finished!</Mesg>
</Complex>
...
</Complexes>
To convert the MM/GBSA output to a better format:
PPL4parseXML --inXML PPL4Track.xml --outXML PPL4Parse.xml
The "pretty" version the MM/GBSA output
<?xml version="1.0" ?>
<Complex>
<!-- Tracking calculation error using XML file -->
<Receptor RecID="2y2vA_A">
<Ligand LigID="1" MinVinaPoseID="1" MinVina="-10.2" MinGBPoseID="2" MinGB="5.2934">
<Pose PoseID="1" Vina="-10.2" GBSA="30.0588">scratch/com/2y2vA_A/gbsa/lig_1/pose_1</Pose>
<Pose PoseID="2" Vina="-10.2" GBSA="5.2934">scratch/com/2y2vA_A/gbsa/lig_1/pose_2</Pose>
<Pose PoseID="3" Vina="-10" GBSA="57.1238">scratch/com/2y2vA_A/gbsa/lig_1/pose_3</Pose>
<Pose PoseID="4" Vina="-9.1" GBSA="17.3861">scratch/com/2y2vA_A/gbsa/lig_1/pose_4</Pose>
<Pose PoseID="5" Vina="-8.9" GBSA="12.672">scratch/com/2y2vA_A/gbsa/lig_1/pose_5</Pose>
</Ligand>
...
</Receptor>
</Complex>
The calculated data are stored under the scratch directory. The directory structure is:
scratch/
com/ --- directory for all complexes
receptor/ --- directory for complex named “receptor” (receptor.pdb)
dock/ --- directory for docking calculation
1/ --- docking ligand 1
2/ --- docking ligand 2
...
gbsa/ --- directory for gbsa calculation
lig_1/ --- ligand 1
pose_1/ --- rescoring pose 1
pose_2/ --- rescoring pose 2
...
lig_2/ --- ligand 2
pose_1/ --- rescoring pose 1
pose_2/ --- rescoring pose 2
...
rec/ --- directory for receptor preparation
...
lig/ --- directory for ligand preparation
1/ --- ligand 1 preparation
2/ --- ligand 2 preparation
...
For the docking complex: the receptor is in scratch/com/receptor/rec/receptor.pdbqt
The ligand is in scratch/com/receptor/dock/1/poses.pdbqt (multiple poses)
For the rescoring comlex : scratch/com/receptor/gbsa/lig_1/pose_1/mmgbsa_results.tar.gz
Unzip the mmgbsa_results.tar.gz. You will find Com_min.pdb
PostProcess MM/GBSA results
<?xml version="1.0" ?>
<Complexes>
<!-- Tracking calculation error using XML file -->
<Complex>
<RecID>1a6q_A_784_minimized_w_metal5</RecID>
<LigID>2</LigID>
<Mesg>Pose 2 has not been completed</Mesg>
</Complex>
<Complex>
<RecID>1a6q_A_784_minimized_w_metal5</RecID>
<LigID>1</LigID>
<Mesg>Finished!</Mesg>
</Complex>
<Complex>
<RecID>1a6q_A_784_minimized_w_metal5</RecID>
<LigID>3</LigID>
<Mesg>Finished!</Mesg>
</Complex>
</Complexes>
Under the scratch/com/<receptor_name>/gbsa, you will find:
lig_1.tar.gz lig_1_checkpoint.txt lig_2 lig_3.tar.gz lig_3_checkpoint.txt
lig_1 and lig_3 are completed so they are tar-zip. The lig_1 and lig_3 are deleted. Lig_2 has not finished so the folder is unchanged.
Note:
1. try to use small number of nodes so that the system is not overwhelming by I/O.
2. Original files under the lig_1, etc will be deleted.
3. PPL4PostProcess can be run while PPL4mmgbsa is running. PPL4PostProcess has restart capability.
The following set up works on Cab/Syrah. For other machines, change the commands accordingly.
use -l python
use python-2.7.10
which pip
pip install --user virtualenv
pip install --user -U virtualenv
pip install --user virtualenvwrapper
export WORKON_HOME=~/.venvs
pip list
pip show virtualenvwrapper
source /g/g92/zhang30/.local/bin/virtualenvwrapper.sh
echo $WORKON_HOME
put the following two command into your .bashrc or .cshrc
export WORKON_HOME=~/.venvs
source /g/g92/zhang30/.local/bin/virtualenvwrapper.sh
mkvirtualenv conveyorlc
git clone [email protected]:LLNL/maestrowf.git
cd maestrowf/
git fetch --all
git checkout -t origin/bugfix/dependency_ordering
pip install -e .
workon conveyorlc
mv run_conveyorlc.yaml ~/W/ConveyorLC/workflow
cd ~/W/ConveyorLC/workflow
maestro run -h
maestro run ./run_conveyorlc.yaml
The "maestro run" will crearte a subdirectory using time stamp (i.e. run_conveyorlc_20180504-155630) To check the job status, run "maestro status"
(conveyorlc)[zhang30@syrah256 workflow]$ maestro status run_conveyorlc_20180504-155630
Step Name Workspace State Run Time Elapsed Time Start Time Submit Time End Time Number Restarts
------------ ------------------------------- ----------------- -------------- -------------- -------------------------- -------------------------- -------------------------- -----------------
PPL4parseXML run_conveyorlc_20180504-155630 State.INITIALIZED --:--:-- --:--:-- -- -- -- 0
PPL1Receptor run_conveyorlc_20180504-155630 State.FINISHED 0:12:01.079873 0:13:01.157428 2018-05-04 15:57:36.256237 2018-05-04 15:56:36.178682 2018-05-04 16:09:37.336110 0
PPL4mmgbsa run_conveyorlc_20180504-155630 State.RUNNING --:--:-- 0:30:03.091956 2018-05-04 16:12:37.699076 2018-05-04 16:11:37.603008 -- 0
PPL2Ligand run_conveyorlc_20180504-155630 State.FINISHED 0:01:00.053917 0:02:00.187536 2018-05-04 15:57:36.256309 2018-05-04 15:56:36.122690 2018-05-04 15:58:36.310226 0
PPL3Docking run_conveyorlc_20180504-155630 State.FINISHED 0:01:00.096606 0:02:00.202646 2018-05-04 16:10:37.473674 2018-05-04 16:09:37.367634 2018-05-04 16:11:37.570280 0
All the results are under the run_conveyorlc_20180504-155630 subdirectory.
description:
name: run_conveyorlc
description: |
Run the conveyorlc code.
env:
variables:
OUTPUT_PATH: ./
labels:
SETUP_MAIN: |
export LBindData=$(CONVEYOR_BIN)/data
export PATH=$(CONVEYOR_BIN)/bin:$(BIN_HOME)/bin:$PATH
export AMBERHOME=$(BIN_HOME)/amber18
export PATH=$AMBERHOME/bin/:$PATH
export INPUTDIR=$(INPUT_DIR)
export WORKDIR=$(OUTPUT_DIR)
dependencies:
paths:
- name: BIN_HOME
path: /usr/gapps/aha/cab
- name: CONVEYOR_BIN
path: /usr/gapps/aha/cab/conveyorlc_test
- name: INPUT_DIR
path: /g/g92/zhang30/W/ConveyorLC/workflow
batch:
type: slurm
host: syrah
bank: bbs
queue: pdebug
study:
- name: PPL1Receptor
description: Run Lingand preparation step of the pipeline.
run:
cmd: |
$(SETUP_MAIN)
$(LAUNCHER)[1n, 3p] PPL1Receptor --input $(INPUT_DIR)/pdb.list --output out
restart: |
$(SETUP_MAIN)
$(LAUNCHER)[1n, 3p] PPL1Receptor --input $(INPUT_DIR)/pdb.list --output out
walltime: "00:30:00"
nodes: 1
procs: 3
- name: PPL2Ligand
description: Run Lingand preparation step of the pipeline.
run:
cmd: |
$(SETUP_MAIN)
$(LAUNCHER)[1n, 16p] PPL2Ligand --sdf $(INPUT_DIR)/pur2.sdf
restart: |
$(SETUP_MAIN)
$(LAUNCHER)[1n, 16p] PPL2Ligand --sdf $(INPUT_DIR)/pur2.sdf
walltime: "00:30:00"
nodes: 1
procs: 16
- name: PPL3Docking
description: Run docking calculations.
run:
cmd: |
$(SETUP_MAIN)
$(LAUNCHER)[1n, 4p] PPL3Docking --recXML $(PPL1Receptor.workspace)/PPL1Track.xml --ligXML $(PPL2Ligand.workspace)/PPL2Track.xml --exhaustiveness 2 --num_modes 1
restart: |
$(SETUP_MAIN)
$(LAUNCHER)[1n, 4p] PPL3Docking --recXML $(PPL1Receptor.workspace)/PPL1Track.xml --ligXML $(PPL2Ligand.workspace)/PPL2Track.xml --exhaustiveness 2 --num_modes 1
depends: [PPL1Receptor, PPL2Ligand]
walltime: "00:30:00"
nodes: 1
procs: 4
cores per task: 4
- name: PPL4mmgbsa
description: Run docking calculations.
run:
cmd: |
$(SETUP_MAIN)
$(LAUNCHER)[1n, 16p] PPL4mmgbsa --comXML $(PPL3Docking.workspace)/PPL3Track.xml
restart: |
$(SETUP_MAIN)
$(LAUNCHER)[1n, 16p] PPL4mmgbsa --comXML $(PPL3Docking.workspace)/PPL3Track.xml
depends: [PPL3Docking]
walltime: "00:30:00"
nodes: 1
procs: 16
- name: PPL4parseXML
description: Convert PPL4Track.xml to readable XML file.
run:
cmd: |
$(SETUP_MAIN)
PPL4parseXML --inXML $(PPL4mmgbsa.workspace)/PPL4TrackTemp.xml --outXML PPL4Parse.xml
depends: [PPL4mmgbsa]
!!!