Lysozyme in Water with aiida-gromacs

This tutorial follows Justin Lemkul’s lysozyme tutorial. We will not explain each individual step as this can be found on Justin’s webpage, but we will link to each page and show the AiiDA equivalant command.

../_images/lemkul.png

Note

The tutorial is written with the assumption that you have a working knowledge of GROMACS. If you are new to GROMACS, we recommend following Justin’s tutorial first to understand the commands and what they do.

Software requirements

For this tutorial, pre-installation of AiiDA, the aiida-gromacs plugin and dependent tools is required, please follow the instructions in the installation section. Here is a brief description of the software used:

  • AiiDA uses a PostgreSQL database to store all data produced and the links between input and output files for each command run. Each submitted command is termed a process in AiiDA.

  • Communication between submitted processes are handled with RabbitMQ and submitted processes are handled with a deamon process that runs in the background.

  • aiida-gromacs requires an installation of GROMACS and the path to where it is installed.

This tutorial also assumes yo have the AiiDA tools running in the background, if not please follow the steps here.

Note

All the files required for this version of the tutorial should be downloaded from our tutorial files and not from the links provided in Justin’s tutorial as slight alterations to these files have been made, and those available via Justin’s tutorial will cause errors if used here.

“AiiDA-fying” Lemkul’s tutorial

  1. We will start from the pdb2gmx step of Justin’s tutorial:

gmx_pdb2gmx -f 1AKI_clean.pdb -ff oplsaa -water spce -o 1AKI_forcefield.gro -p 1AKI_topology.top -i 1AKI_restraints.itp

Note

There may be slight differences in commands between the tutorial and that by Justin Lemkul, this is simply down to the way we are recording provenance, which requires non-interactive input into the GROMACS tools.

Note

After each of the steps you should run verdi to view the status of the submitted process before moving onto the next step, you do this by running:

verdi process list -a

A successfully finished process will exit with code [0].

  1. Next we will create the box and solvate

    Firstly the box:

    gmx_editconf -f 1AKI_forcefield.gro -center 0 -d 1.0 -bt cubic -o 1AKI_newbox.gro

    Then solvate:

    gmx_solvate -cp 1AKI_newbox.gro -cs spc216.gro -p 1AKI_topology.top -o 1AKI_solvated.gro

  2. Add ions

    Firstly we will use the grompp preprocessor:

    gmx_grompp -f ions.mdp -c 1AKI_solvated.gro -p 1AKI_topology.top -o 1AKI_ions.tpr

    Followed by genion:

    gmx_genion -s 1AKI_ions.tpr -p 1AKI_topology.top -pname NA -nname CL -neutral true -o 1AKI_solvated_ions.gro

  3. Then minimise the structure

    Firstly we will use grompp:

    gmx_grompp -f min.mdp -c 1AKI_solvated_ions.gro -p 1AKI_topology.top -o 1AKI_minimised.tpr

    Then mdrun to minimise:

    gmx_mdrun -s 1AKI_minimised.tpr -c 1AKI_minimised.gro -e 1AKI_minimised.edr -g 1AKI_minimised.log -o 1AKI_minimised.trr

  4. Now we will equilibrate with NVT

    Firstly we will use grompp:

    gmx_grompp -f nvt.mdp -c 1AKI_minimised.gro -r 1AKI_minimised.gro -p 1AKI_topology.top -o 1AKI_nvt.tpr

    Then mdrun to equilibrate NVT:

    gmx_mdrun -s 1AKI_nvt.tpr -c 1AKI_nvt.gro -e 1AKI_nvt.edr -g 1AKI_nvt.log -cpo 1AKI_nvt.cpt -o 1AKI_nvt.trr

  5. Followed by equilibration with NPT

    Firstly we will use grompp:

    gmx_grompp -f npt.mdp -c 1AKI_nvt.gro -r 1AKI_nvt.gro -t 1AKI_nvt.cpt -p 1AKI_topology.top -o 1AKI_npt.tpr

    Then mdrun to equilibrate NPT:

    gmx_mdrun -s 1AKI_npt.tpr -c 1AKI_npt.gro -e 1AKI_npt.edr -g 1AKI_npt.log -cpo 1AKI_npt.cpt -o 1AKI_npt.trr

  6. We are now ready for production MD.

    Firstly we will use grompp:

    gmx_grompp -f prod.mdp -c 1AKI_npt.gro -t 1AKI_npt.cpt -p 1AKI_topology.top -o 1AKI_prod.tpr

    Then mdrun for production run:

    gmx_mdrun -s 1AKI_prod.tpr -c 1AKI_production.gro -e 1AKI_production.edr -g 1AKI_production.log -o 1AKI_production.trr

    If running on GPU then something like:

    gmx_mdrun -s 1AKI_prod.tpr -c 1AKI_production.gro -e 1AKI_production.edr -g 1AKI_production.log -o 1AKI_production.trr -bonded gpu -nb gpu -pme gpu -ntmpi 1 -ntomp 5 -pin on

That is it! You’ve ran your first GROMACS simulation with AiiDA.

Note

The majority of the commands used in Justin’s tutorial have an equivalent in the aiida-gromacs plugin. To view all gmx commands available in the plugin, run:

verdi plugin list aiida.calculations

Anything starting with gromacs. is available in the plugin. To use other commands not available in the plugin, you can use the genericMD CLI, which allows you to save any command you want to keep track of with AiiDA.

Viewing and sharing data

You can now view the provenance graph of the simulation by running:

verdi node graph generate <PK>

Where <PK> is the process ID of the last process you ran.

Note

The provenance graph will show all the steps you’ve taken in the simulation, and the connections between the input and output files for each step. This is a great way to visualise and keep track of your simulations.

The simulation steps can also be viewed on the terminal by running:

verdi data provenance show

The provenance can also be archived for sharing with others, to do this run:

verdi archive create --all archive_name.aiida

Where --all saves all the data in the AiiDA profile.

Now, have a go at exploring the AiiDA archive file on a demo database application called BioSimDB. This is a web-based application that allows you to view and store your simulation data provenance in a local database.