Implementing complex CVs in PLUMED

[Massimo Bocus]

Dear PLUMED users/developers,

I'd like to try out some simulations (PLUMED coupled with CP2K on a supercomputing architecture) with a collective variable that would look something like a coordination number, in the sense that I'd need to iterate over distances  between pairs of atoms belonging to two groups. 

Now, the two groups are quite large in my system, so the first solution I thought about would be to use a Python script to write an extremely long PLUMED input (combining DISTANCE, MATHEVAL and so on). This feels however like a very 'brute force' solution, therefore I would like to know if there are possible alternatives to it (for this case but also in general for the implementation of complex CVs without directly modifying the code of PLUMED).

I haven't found anything by looking online or in this group, so any help is much appreciated.

Thanks in advance, best regards,

Massimo

Center for Molecular Modeling - Ghent University

[Gareth Tribello]

Use a MiultiColvar.  To get started look here:

https://www.plumed.org/doc-v2.7/developer-doc/html/_adding_a_multi_colvar.html

Better still do this tutorial:

https://www.plumed.org/doc-v2.7/user-doc/html/julich-devel.html

You can adapt what it is telling you to do to your particular purposes

Gareth

--
You received this message because you are subscribed to the Google Groups "PLUMED users" group.
To unsubscribe from this group and stop receiving emails from it, send an email to plumed-users...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/plumed-users/9f236687-12a1-4081-9f7b-4b4742d054ebn%40googlegroups.com.

[Giovanni Bussi]

If it's a sum of a function applied to pairwise distances, you can use COORDINATION with a custom function without implementing anything.

If it's not a sum, search among the available multicolvars if one of them is suitable for your need, or modify src/colvar/CoordinationBase.cpp, or implement a new multicolvar.

In general, if you are using CP2K for ab initio MD I suspect that even defining a very long plumed input should not significantly slow down your simulation.

Giovanni

--