Covalent Library Preparation 2024: Difference between revisions

From DISI
Jump to navigation Jump to search
(Created page with "Modify warhead to covalent adduct with SiH3 added. == Prepare “SMILES ID” file. == ssh n-1-17 (or another development node) Source environment with RDkit source /mnt/nfs/ex9/work/ttummino/miniconda/etc/profile.d/conda.sh conda activate base3.7 python ~ak87/PROGRAM/convert_smiles_to_covalent.py ald bbv-cov-ald.smi bbv-cov-ald-test.smi This script enumerates all stereoisomers and converts SMILES to the covalent ones. Currently, only conversions for aldehydes...")
 
No edit summary
Line 34: Line 34:
  csh ~elisfink/scripts/09-2022-ligbuild/0001_wrapper_queue_build_smiles_ligand_mod_corina_covalent.csh input.smi
  csh ~elisfink/scripts/09-2022-ligbuild/0001_wrapper_queue_build_smiles_ligand_mod_corina_covalent.csh input.smi


== Updated version from Khanh ==
Note: I tried following the pipeline above. Although it worked great on gimel (a CentOS server), I have trouble running it on CentOS 7 servers due to AMSOL that it used which was compiled on a very out-of-date fortran complier. So I recompiled AMSOL and refactored the build3d pipeline to use Python3 (I know, another copy of DOCK that is untracked)
=== Usage ===
This script only works on servers with GCC version 4+ (slurm nodes) and it won't work on our sge nodes running on CentOS 6
Example:
ssh n-1-17
source /nfs/ex7/blaster/templates/build3d_mol2/dock37_pipeline/env.sh
export SOURCE_FILE=input.smi
// build covalent molecules
bash $DOCKBASE/ligand/generate/build_database_ligand.sh --no-db --no-solv --no-mol2 --single --covalent -H 7.4
// build db2 from mol2
bash /nfs/ex7/blaster/templates/build3d_mol2/dock37_pipeline/build_ligand.sh input.mol2 --name="Name" --smiles="SMILES"


[[Category:Covalent]]
[[Category:Covalent]]

Revision as of 22:35, 3 May 2024

Modify warhead to covalent adduct with SiH3 added.

Prepare “SMILES ID” file.

ssh n-1-17 (or another development node)

Source environment with RDkit

source /mnt/nfs/ex9/work/ttummino/miniconda/etc/profile.d/conda.sh
conda activate base3.7
python ~ak87/PROGRAM/convert_smiles_to_covalent.py ald bbv-cov-ald.smi bbv-cov-ald-test.smi

This script enumerates all stereoisomers and converts SMILES to the covalent ones. Currently, only conversions for aldehydes and nitriles is supported. Usage:

  • First arg: ald or nitr for aldehyde or nitrile
  • Second arg: input smi file
  • Third arg: output file name


Preparing ligands

SSH Gimel

ssh gimel

Set to csh

csh

Source Environments

source ~elisfink/.csh_corina
setenv DOCKBASE /nfs/soft/dock/versions/dock37/DOCK-3.7-trunk
source /nfs/soft/dock/versions/dock37/DOCK-3.7-trunk/env.csh

Run Script with input.smi as your prepared SMILES file

csh ~elisfink/scripts/09-2022-ligbuild/0001_wrapper_queue_build_smiles_ligand_mod_corina_covalent.csh input.smi

Updated version from Khanh

Note: I tried following the pipeline above. Although it worked great on gimel (a CentOS server), I have trouble running it on CentOS 7 servers due to AMSOL that it used which was compiled on a very out-of-date fortran complier. So I recompiled AMSOL and refactored the build3d pipeline to use Python3 (I know, another copy of DOCK that is untracked)

Usage

This script only works on servers with GCC version 4+ (slurm nodes) and it won't work on our sge nodes running on CentOS 6 Example:

ssh n-1-17
source /nfs/ex7/blaster/templates/build3d_mol2/dock37_pipeline/env.sh
export SOURCE_FILE=input.smi
// build covalent molecules
bash $DOCKBASE/ligand/generate/build_database_ligand.sh --no-db --no-solv --no-mol2 --single --covalent -H 7.4
// build db2 from mol2
bash /nfs/ex7/blaster/templates/build3d_mol2/dock37_pipeline/build_ligand.sh input.mol2 --name="Name" --smiles="SMILES"