Dear Adam,
Thanks for making MDWeb available. Such an excellent service to the research community!
I am now trying to learn it so that I can use it to generate a RMSd plot for my docked compound.(Please forgive me if my question is silly - I should declare that I have zero knowledge/experience in MD. )
I could follow the steps in your tutorial up to the point where I downloaded a f01.moby.tgz file.
I would like to use MDWeb, not Gromacs, for analysis. So I guess the next step is to start a new project and choose Analysis (MD Trajectory). Base trajectory setting is where I started to get confused/lost.
(1)
Unzipped f01.moby.tgz does not contain an XTC file. Your tutorial has some info about XTC trajectory in the section “3. Uploading a trajectory”. Is that how I can create an XTC file? It seems to say that I should install Gromacs first. So I have installed Gromacs on my win10 laptop and I tried typing in trjconv… but it said “command not found”.
(2)
Also, there are multiple itp files in my unzipped f01.moby.tgz. Which should I upload? Or how should I determine which to upload?
I am stuck at this point. Could you help?
Thanks and have a good day!
Chai
Dear Chai,
thanks for using our tool and for your kind words!
MDWeb server offers an easy way to setup your structure and obtain a system (molecule with hydrogen atoms, surrounded by water and counterions, energetically minimized and equilibrated) ready to be used as input for a production MD. But it is not running the simulation, it is just preparing the system. This is of course because of the need of more computational resources and disk space. So the Protein MD Setup tutorial you followed has generated the files you need to start the MD: topology, coordinates and velocities, but with no associated trajectory (XTC). As you chose the Gromacs workflow, you should use gromacs to run the simulation and obtain the trajectory. After running the simulation (in your local machine), you will then have the possibility to upload a short trajectory to MDWeb and run some quality control analysis such as RMSd or Radius of Gyration, and also some flexibility analysis with the link to FlexServ server.
Now, to run the simulation in your local machine, I recommend you to take a look at some Gromacs tutorials. There is a brand new gromacs tutorials page here. I guess trjconv is not found in your system because the last versions of gromacs only recognise the gmx or gmx_mpi binary, followed by the name of the tool (e.g. gmx trjconv).
For the itp files, the main topology file in gromacs is the .top file, which has different include lines inside, linking all the itp files. So the one to be used is the .top file.
All that said, I would recommend you to take a look at the BioExcel Building Blocks (biobb) website, which is a more recent library that can help you on the MD setup and run processes. You could try to run a setup with your structure using the Protein MD Setup workflow here, and then add an extra step to run the simulation.
Hope this helps!
Regards,
-Adam-
Dear Adam,
Thank you very much for taking your time to give me such a detailed explanation.
I think I understand better now about MDWeb.
So, how about biobb? Will it allow me to perform the whole process in it?
Or like MDWeb, I still need to use Gromacs, and then only return to biobb?
Thanks,
Chai
Dear Chai,
BioBB will help you in the process of setup and run a MD simulation using gromacs, but easing the installation process. You have different demonstration workflows available as Jupyter Notebooks, with very easy instructions about how to install and run them in your own computer. You can also open them in our BioExcel Binder to play and familiarise with them without any installation process needed, but I recommend you to try a local installation. If you are using Windows, this tutorial may help you.
Please take a look at the Protein MD Setup and the Protein-Complex (with Ligand) MD Setup workflows to see how BioBB workflows work.
There’s a specific section in this ask.bioexcel forum related to the BioBB library where the main developers will help you with any question/doubts you might have.
Thanks again for your interest in our tools!
Best regards,
-Adam-
Dear Adam,
Thanks a lot for your very detailed guidance.
My impression is that BioBB is a more user-friendly tool than MDWeb, yes?
Looks like there will be a lot to learn. But thankfully, now I know who and where to ask questions later on
I will try my best!
Thanks,
Chai