Contribution of project Extend MDAnalysis Interoperability

[Purnawan P]

Dear mentor,

My name is Purnawan Pontana Putra, and I am a researcher at Universitas Andalas. I am interested in participating in the Summer Code Mdanalysis, particularly in Project 2, which is to Extend MDAnalysis Interoperability. I have prior experience with several molecular dynamics software, such as GROMACS, AMBER, OPENMM, and NAMD. I am also proficient in using Python-based software, including pytraj, parmed, cpptraj, and Mdanalysis.

I am excited about the opportunity to participate in this project as I believe it will enhance my knowledge and skills, particularly in the context of my work at the university. I am eager to collaborate with other participants and the project mentors to achieve the project's goals.

Thank you for considering my application.

Best Regards,
Purnawan Pontana Putra

[Rocco Meli]

Hi Purnawan Pontana Putra,

Welcome to MDAnalysis!

We are happy to hear you are interested to participate to GSoC, and that you already have prior experience with MD. We are curious to hear, how have you used MDAnalysis in your previous work?

Project 2 is an interesting and important project if you work in the MD field. A good point to start would be to have a look to the "Interoperability Roadmap", if you haven't already:

https://www.mdanalysis.org/2020/08/03/roadmap/

 Is there a particular package you would like to have interoperability with?

Since you are already familiar with MDAnalysis, you can look at the sections of the User Guide explaining how to contribute:

https://userguide.mdanalysis.org/stable/contributing.html

As per the MDA GSoC FAQ (https://github.com/MDAnalysis/mdanalysis/wiki/GSoC-FAQ), candidates must have code merged to be eligible to apply to GSoC with MDAnalysis (see https://github.com/MDAnalysis/mdanalysis/wiki/GSoC-FAQ#what-should-i-do-for-a-successful-proposal). We look forward to your contribution.

Best,

Rocco

[Purnawan P]

I am interested in analyzing the molecular dynamics simulation results from the xTB software package https://github.com/grimme-lab/xtb using MDanalysis. I want to analyze it quickly, such as seeing energy stability and molecular motion during the simulation. Here's my github address: https://github.com/purnawanpp

[Oliver Beckstein]

Hello Purnawan Pontana Putra,

Extending MDAnalysis into the qm/materials space is interesting. I don’t know xtb, though, so the following is just a very rough first take on your idea; ultimately, your proposal has to convince us that it’s a worthwhile project and that you’re able to successfully execute it.

If you want to work with xtb output files then you should look into topology reading and trajectory reading (and possibly auxiliary files). 

If you want to work with the xtb-python API then you should look into our converter framework.

In both cases, you should look into potentially making it an MDAKit (or at least keeping this option, depending on how involved installation of xtb is; they have a conda-forge package but I don’t know what dependencies it pulls in). 

Oliver

--
GSoC 2023: Read https://www.mdanalysis.org/2023/02/22/gsoc2023/
Our FAQ: https://github.com/MDAnalysis/mdanalysis/wiki/GSoC-FAQ
---
You received this message because you are subscribed to the Google Groups "MDAnalysis Google Summer of Code" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gsoc+uns...@mdanalysis.org.
To view this discussion on the web visit https://groups.google.com/a/mdanalysis.org/d/msgid/gsoc/da983871-9d28-4428-8b43-763f9695b67cn%40mdanalysis.org.

--

Oliver Beckstein (he/his/him)

email: orbe...@mdanalysis.org
twitter: @orbeckst

GitHub: @orbeckst

MDAnalysis – a NumFOCUS fiscally sponsored project

https://www.mdanalysis.org/

[Purnawan P]

Thanks for the opinion and advice. I have successfully run it like analyzing trajectory data and calculating RMSD from xtb.trj using this script https://github.com/purnawanpp/xtb_md/blob/main/xtb_md.ipynb and https://github.com/purnawanpp/xtb_md/blob/main/rmsd.ipynb

What do you think if implemented in MDAnalysis?

On Tuesday, 4 April 2023 at 00:24:28 UTC+7 Purnawan P wrote:

Thanks for the opinion and advice. I have successfully run it like analyzing trajectory data and calculating RMSD from xtb.trj using this script https://github.com/purnawanpp/xtb_md/blob/main/xtb_md.ipynb and https://github.com/purnawanpp/xtb_md/ blob/main/rmsd. ipynb. 
What do you think if implemented in MDAnalysis?

[Oliver Beckstein]

Hello Purnawan Pontana Putra,

It looks that you’re off to a good start in that you have an idea how to get the information from the trajectories but from what you showed I am not sure yet how much would be needed to be done in MDA. For instance, do you need to write a new coordinate reader or can you use the existing XYZ reader. Do you want to read energies from a separate file and then associate it with the trajectory using the auxiliaries framework?

There should be enough work to justify at least a medium size (175h) GSoC project.

Ultimately, your proposal has to convince us that this is a worthwhile project, see https://github.com/MDAnalysis/mdanalysis/wiki/GSoC-FAQ#what-are-your-evaluation-criteria . I’d expect to see a discussion of how your proposed code fits into MDAnalysis’s standard trajectory API (and how you’ll read the topology information), how you’ll process energies (as auxiliaries or as part of the trajectory reading?), and possibly a discussion about API-level interoperability with xbt via xbt-python (or why you’d decide to NOT work on API interoperability). Do not forget to include to discuss (and allocate time!) for testing and documentation.

Given that the submission deadline is TOMORROW you should really start writing. Unfortunately, you’re quite late so there might not be a lot of opportunity to give detailed feedback on the proposal before the deadline.

Best,
OIiver