ZINC-22:3D Building

From DISI
Revision as of 18:25, 30 June 2020 by Btingle (talk | contribs) (→‎Usage)
Jump to navigation Jump to search
The printable version is no longer supported and may have rendering errors. Please update your browser bookmarks and please use the default browser print function instead.

Introduction

You can request access to the zinc 3d build repository here:

https://github.com/btingle/zinc-3d-build

The contents of this github can be found on /nfs/home/xyz/btingle/zinc_deploy/zinc-3d-build. There are two top-level bash scripts in this repository that should be used for the 3d building process:

  • build_3d_smiles.bash
  • export_zinc_ids.bash

export_zinc_ids.bash

This script exports the SMILES for all substances from a chosen TIN database and annotates them with unique ZINC codes.

Example Usage

export ZINC_PORT=5434
export EXPORT_DEST=/local2/3d/export
./export_zinc_ids.bash

ZINC_PORT is the port number of the database on this machine that should be exported from.

EXPORT_DEST is where the annotated smiles/ZINC ID file should be saved.

The name of the output file is formatted as follows:

$EXPORT_DEST/[PARTITION_RANGE]_[PORT].export

build_3d_smiles.bash

This script submits jobs on the cluster to build the 3d data for the output of export_zinc_ids.bash

Example Usage

export SMILES_FILE=/local2/3d/export/H30P400_H30P400_5434.export
export TEMPDIR=/local2/3d/build/scratch
./build_3d_smiles.bash

The nodes allocated for these jobs are defined in nodelist.txt.

The output of this script will be in $TEMPDIR/[SMILES].out split into a number of batches. Each batch .tar file contains the output of 5000 jobs. Each job outputs a .tar.gz file containing the results of 3d building for 20 substances.