GRRM
The global reaction route mapping GRRM® program is a commercial software developed for the analysis of chemical reaction pathways. This program is capable of exhaustively exploring the potential energy surface of a collection of atoms, automatically finding molecular geometries of reactants, transition states, and products. GRRM23 works in conjunction with quantum chemistry software such as, but not limited to, Gaussian 16.
License Agreement¶
To obtain a license (and pricing), please submit a request at https://global.hpc.co.jp/contact/. Alternatively, contact Kenshiro Kimura for GRRM23 application instructions. GRRM23 uses a floating license approach for validating users; it requires the setup of a license server. For this, HPC SYSTEMS Inc., the company that sells GRRM23, allows users to connect to license servers at their company site in Japan at no additional costs. Alternatively, a license server can be set up at the user’s university in Canada; for this option, please contact technical support at HPC SYSTEMS Inc.
Installation Instructions¶
After buying a license, users will receive a link and instructions to download a compressed tar.bz2 file. Download the file and follow the instructions below (substitute the path to your actual home directory where /home/user is mentioned).
In your home directory, create a directory for the installation and another one for the executables:
Move the compressed GRRM23 file to $HOME/GRRM23_Installation and extract it there:
The extraction of the tar.bz2 should generate an installation script (install.sh); from /home/user/GRRM23_Installation, execute the script as follows:
The installation will display the following configuration summary:
Installation configuration:
GRRMroot = /home/user/GRRM23_RUN
Intel MPI installation prefix = /home/user/GRRM23_RUN/mpi-rt
Are you sure to start installation? [Y/n]
Type “Y” and press “Enter” to proceed with the installation. Once the installation is finished, a “GRRM INSTALLATION DONE” message should appear:
Installing Intel MPI ... done successfully.
Installing GRRM binary ...done successfully.
Installing STLib ...done successfully.
GRRM INSTALLATION DONE.
IMPORTANT[1]: Before using GRRM, please insert lines below to your shell (/bin/bash) configuration file.
IMPORTANT[2]: Please consult your system administrator about the STLM server IP address and port number.
source /home/user/GRRM23_RUN/mpi-rt/mpi/2021.6.0/env/vars.sh
export PATH="/home/user/GRRM23_RUN:${PATH}"
export LD_LIBRARY_PATH="/home/user/GRRM23_RUN:${LD_LIBRARY_PATH}"
export GRRMroot="/home/user/GRRM23_RUN"
export subgrr="GRRM23.out"
export subgau=""
export subchk=""
export subuchk=""
export submol=""
export subsiesta=""
export submpi="mpirun"
export I_MPI_FABRICS="shm:tcp"
export I_MPI_FALLBACK=0
export I_MPI_HYDRA_BOOTSTRAP="ssh"
export STLM_SERVER_ADDRESS="XXX.XXX.XXX.XXX"
export STLM_SERVER_PORT=XXXXX
Configuration File Update
Before using GRRM, please insert the lines provided in the installation output into your shell (/bin/bash) configuration file (e.g., ~/.bashrc).
STLM Server Details
Please consult your system administrator about the STLM server IP address and port number. This information is crucial for license validation.
Copy the last lines (from “source” to the last “export” line) into your .bashrc in your home directory. Since G16 is not loaded on the Alliance's login nodes, you need to manually add g16, formchk, and unfchk into the corresponding lines below (subgau, subchk, and subuchk, respectively). For validating your license, please replace the XXX values at STLM_SERVER_ADDRESS and STLM_SERVER_PORT with the information provided by HPC SYSTEMS Inc. for your license (HPC SYSTEMS Inc. will send users this information via email):
source /home/user/GRRM23_RUN/mpi-rt/mpi/2021.6.0/env/vars.sh
export PATH="/home/user/GRRM23_RUN:${PATH}"
export LD_LIBRARY_PATH="/home/user/GRRM23_RUN:${LD_LIBRARY_PATH}"
export GRRMroot="/home/user/GRRM23_RUN"
export subgrr="GRRM23.out"
export subgau="g16"
export subchk="formchk"
export subuchk="unfchk"
export submol=""
export subsiesta=""
export submpi="mpirun"
export I_MPI_FABRICS="shm:tcp"
export I_MPI_FALLBACK=0
export I_MPI_HYDRA_BOOTSTRAP="ssh"
export STLM_SERVER_ADDRESS="XXX.XXX.XXX.XXX"
export STLM_SERVER_PORT=XXXXX
Source your .bashrc file and logout:
After login, type GRRM23p on the terminal. If the installation was successful, the following text should appear:
Fir and Nibi have G16 installed for their users. Compute nodes on both systems can connect to the Internet for the purpose of connecting to any required license servers.
Running GRRM¶
For input examples, please refer to the GRRM23 user manual. In general, GRRM23 requires only one input file (Job.com) with the Cartesian coordinates of a molecule plus methodology options; in some cases, GRRM23 will require the output of a previous GRRM23 job.
Sample Slurm Script for a Single Node¶
For a single node (-N 1) and 12 cores (-n 12), running from the user’s /scratch directory:
#!/bin/bash
#SBATCH -t 1-00:00
#SBATCH -J Job_SLURM
#SBATCH -o Job.SLURMout
#SBATCH -e Job.SLURMerr
#SBATCH -N 1
#SBATCH -n 12
#SBATCH --mem-per-cpu=4000M
#SBATCH --account=def-users_account
input=Job
cd /home/user/scratch/
module load gaussian/g16.c01
export GAUSS_SCRDIR=/home/user/scratch/
GRRM23p ${input} -p1 -s80000
-p1: number of parallel processes;-s80000: signals when to create a checkpoint file (in seconds); you must tell GRRM23 when to create the checkpoint files so that an interrupted/killed job can be restarted properly.
Example Using ORCA¶
#!/bin/bash
#SBATCH -t 0-18:00
#SBATCH -J glucose_SLURM
#SBATCH -o glucose.SLURMout
#SBATCH -e glucose.SLURMerr
#SBATCH -N 1
#SBATCH -n 16
#SBATCH --mem-per-cpu=4000M
#SBATCH --account=def-users_account
input=glucose
cd /home/user/scratch/Pyrolysis_of_Models/glucose
module load StdEnv/2023 gcc/12.3 openmpi/4.1.5
module load orca/6.0.0
GRRM23p ${input} -p1 -h15
Sample Slurm Script for Multiple Nodes¶
At this moment, since we are using GRRM23 with G16 primarily, we can only use a single node, as per Gaussian 16 license policies.
External Resources¶
Acknowledgements¶
This documentation was contributed by Eduardo Romero Montalvo.