With my friends, I am going to write a script for GROMACS to make running simulations automatic. It will take a pdb file as an input and ask some questions. Elese will be done automatically. I think it would be good for people who want to run a simulation but do not know GROMACS.
I want to hear your opinion about this idea. Will it be useful? Should it ask for everything, or rely on some default mdp files? What simulation types should we include?
Of course you should know what you are doing (to some extent) to use a tool properly. But that does not mean that you have to make it so hard that only the experts can use it. By that argument you should only use, say an NMR if you completely understand every single aspect of the instrument (incl folding, off resonance effects, resolution, processing etc etc). For BASIC work you can use standard parameters and that gets acceptable results in many cases.
There is another aspect to this. I teach molecular modeling.... I would love to use AMBER or GROMACS since this (AMBER) is what we are using for research, however a script driven program like sander is not really suitable for teaching in a classroom setting. Which means we use Hyperchem or Sybyl to look at basic dynamics....
This would be very interesting and useful project to bring molecular simulation to non-experts. For example, the quantum chemistry program Gaussian is very popular because many calculations are close to being "black box" calculations. Since defaults are defined that work in the vast majority of the cases the user does not need to worry about the mechanics of the calculation in most cases. I don't know if such a thing would be possible with GROMACS but if common simulations could be partially automated it would be helpful to many people.
Hello.
That is an excelent idea. Unfortunately, some errors or glitches that may occur in the whole process may still require a more expert intervention from the user. If i'm not mistaken, in GROMACS webpage you have some GUI examples that weren't much used sue to this issue.
Best regards,
Ricardo Ferreira
Terrible (IMHO) idea. When people do the things without any understanding of what they are doing, "very bad things" may happen.
GROMACS is a pretty simple package to use, and this often leads to publications when authors have no idea of what they have done. I have been a reviewer of manuscripts where "scientists", for instance, placed a protein molecule into the periodic box whose dimensions were smaller that the molecule itself and presented an RMSD plot with values reaching huge values within picoseconds (and were also saying that the protein was stable over the course of simulation).
“Make things as simple as possible, but not simpler." (A. Einstein)
As for the Gaussian popularity, I don't personally think it is due to its being a "black-box" like. There are not many software packages that can do similar things with a comparable efficiency, and that also explains why people agree to pay quite a bit of $ instead of using other (free) programs.
It is not just possible, automation for Gromacs is already achieved. Gromacs runs fully automatically on WeNMR portal. You can run it simply from a structure file or from a binary input file. But you need to get access to WeNMR Grid. Check out the links:
http://www.wenmr.eu/wenmr/md-simulations-using-gromacs-web-server
http://haddock.science.uu.nl/enmr/services/GROMACS/main.php
http://haddock.science.uu.nl/enmr/services/GROMACS/server_full.php
http://haddock.science.uu.nl/enmr/services/GROMACS/server_partial.php
You may also try DESMOND which is a very elegant MD software. It runs on Schrodinger Maestro molecular modeling environment. It has a very user-friendly GUI. It is much faster than Gromacs and above all it's FREE and OPEN SOURCE. It is provided by D.E. Shaw Research Group.
http://www.schrodinger.com/productpage/14/3/
http://www.deshawresearch.com/
A very good package to handle GROMACS jobs is Copernicus: http://copernicus-computing.org/
It's ready. I did it with fellows from my Students Scientific Association. Here is a paper for our students' journal (sory it's in polish). Please give me constructive criticism so I'll make it better
https://www.researchgate.net/publication/248398973_MYGENerator__a_tool_for_simplified_and_automatized_usage_of_the_GROMACS_package?ev=prf_pub
Article MYGENerator – a tool for simplified and automatized usage of...
I think this is a very bad idea. One should know what he/she is doing. Every package has its limitation. If one uses it as a black box he/she will not be able to understand its weak points.
Of course. But it is something similar to some wet lab kits that got eg. "buffer A" "buffer B" etc and You use it because it works. It surely isn't something You can build a paper on, but it can help just to "see" flexibility of peptide (what my colleague is actually doing now)
Of course you should know what you are doing (to some extent) to use a tool properly. But that does not mean that you have to make it so hard that only the experts can use it. By that argument you should only use, say an NMR if you completely understand every single aspect of the instrument (incl folding, off resonance effects, resolution, processing etc etc). For BASIC work you can use standard parameters and that gets acceptable results in many cases.
There is another aspect to this. I teach molecular modeling.... I would love to use AMBER or GROMACS since this (AMBER) is what we are using for research, however a script driven program like sander is not really suitable for teaching in a classroom setting. Which means we use Hyperchem or Sybyl to look at basic dynamics....
Guy, for organizing a large number of GROMACS simulations there's already a package called Copernicus.
- Badges
- Science topic