I have taken The Bioinformatics Resources & Applications Facility (BRAF) at C-DAC. But donot know how to use it for MD simulation of protein-ligand complex using Gromacs program. Could anyone tell me the step by step procedure to perform MD simulation of protein-ligand complex using this server facility? and how to use environment module to use different version of gromacs in braf server??
1.LOGIN to your server
2.install GROMACS
3. Follow tutorials: http://www.bevanlab.biochem.vt.edu/Pages/Personal/justin/gmx-tutorials/complex/index.html
to familiarize the software.
4.Explore the options at every step, and apply appropriate ones according to your experiment. Try running for different number of steps
5. Analyse the result carefully by scanning the log, most importantly apply some basic rules of biology(common sense).
Hello Prakash Jha
Do not run jobs directly on qlogin. Prepare the .tpr file in your local system and use the .tpr & .pbs file to perform the jobs.
Thank you Thirumal Kumar.. I know the basic steps of gromacs..
I want to know how to perform it on server??
Upto which steps I have to perform on local system..?
I have done the nvt npt simulation in my local computer..
Then there is last step of simulation:
grompp -f md.mdp -c npt.gro -t npt.cpt -p topol.top -o md.tpr
mdrun -deffnm md
These last two steps i have to perform on server??
Which tpr file u are saying about??
and how to use pbs file..
Please can you tell me??
Thanks in advance
If you've finished nvt and npt equilibration then you're all set to grompp and do mdrun.
It will depend on what gromacs version you have locally on your desktop and what is available on the server. The *.tpr files are not cross compatible with other versions.
So, if you have GROMACS 2018 locally but GROMACS 2016 is on the cluster, then "Yes" you'll have to "grompp" on the server AND do the "mdrun" for your production run. But, if both your local machine and the server have the same version of gromacs then, you can "grompp" on your local machine and then do "mdrun" on the server.
If it's anything like here though, you'll probably need to write a bash script or something that allows for restarts and prevents overwriting your data, as you probably won't get uninterrupted access to the server. That said, I don't know how that server works. I'd definitely recommend asking around for scripts from other people that use that server specifically.
If you know how to use the software and run a simulation, RG's members cannot help you, unless someone happen to work at your site.
What you need is to ask the system administrator of your site the next three questions:
1. If GROMACS is installed, and if so, how to load or activate the environment to run it.
2. How to submit a job to the job queue of you system.
3. How to select from the available resources those accessible to you.
In addition, MDRUN is usually pretty automatic, but if it cannot guess correctly how many CPU cores/GPU cards are available, if may fail to run or run very inefficiently.
You may need to learn how to inform GROMACS of the site's HARDWARE topology and may need to run a few simulation trials using different amount of resources to help you decide how to exploit your computer site with GROMACS at optimum efficiency.
Usually you may run a series of short trial with a test MD system and using different combinations of cores/GPU's to find how the wall time scales with increasing resources and how the ns of simulation relate to the effetive CPU/GPU time used. This cannot be generalized, it has to be tested for your case at your site.
Best wishes,
Rogelio
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Bioinformatics