Project

General

Profile

Bug #490

default index groups duplicated, non-SOL removed

Added by Peter Kasson over 9 years ago. Updated over 9 years ago.

Status:
Closed
Priority:
Normal
Assignee:
Erik Lindahl
Category:
mdrun
Target version:
Affected version - extra info:
Affected version:
Difficulty:
uncategorized
Close

Description

In the lead-up to the 4.5 beta, some changes were made to the default index groups. SOL is now replaced with Water, which I can deal with. But the default groups seem replicated several times over, and the non-SOL or equivalent default group has disappeared.

e.g.
0 System : 1178714 atoms
1 Protein : 0 atoms
2 POPE : 94608 atoms
3 POPC : 102 atoms
4 LN6 : 8 atoms
5 DNA : 0 atoms
6 POPE : 94608 atoms
7 POPC : 102 atoms
8 LN6 : 8 atoms
9 RNA : 0 atoms
10 POPE : 94608 atoms
11 POPC : 102 atoms
12 LN6 : 8 atoms
13 Water : 1083996 atoms
14 POPE : 94608 atoms
15 POPC : 102 atoms
16 LN6 : 8 atoms

(for a system containing POPE, POPC, LN6, and Water)

History

#1 Updated by Erik Lindahl over 9 years ago

This has already been fixed in git as far as I know (check beta 2 or the release-4-5-patches branch). I'm tentatively marking it as 'fixed' for that reason, but we can continue the discussion.

SOL is still there as an alias to "Water" since it is used by a number of mdp files. I'm not sure if Non-SOL is equally important, but I'm open to arguments.

We'd like to avoid having too many default groups, and the new residuetypes.dat file both enables is to generate some useful default groups like "Water_and_ions", and you can easily define a new "lipid" group by listing a set of molecules with a new arbitrary type in residuetypes.dat

Cheers,

Erik

#2 Updated by Peter Kasson over 9 years ago

Fair enough; it would be really useful to have non-Water or /non-SOL as a default group (also I used to script for the creation of such a group, but I changed over to use the default option. Having everything suddenly change breaks files that used to work...)

We could debate whether to automatically lump ions in with water too.

#3 Updated by Erik Lindahl over 9 years ago

non-Water is present in the newer versions - it was just a bug that it didn't appear at first.

I just fixed another bug that gave rise to double copies of protein groups when you had two protein segments. It's been committed as hash d688115235f66a6fa2df762b5f56d82af9bfcf59.

I think it's dangerous to have a group "Water" that really contains non-water-residues too. For a while I considered having a "Solvent" group, but that too is problematic. A novice user could expect that group to contain e.g. his urea molecules, methanol, or whatever other non-water solvent molecules s/he might have. With "Water_and_ions" it is at least obvious what you get!

Cheers,

Erik

#4 Updated by Peter Kasson over 9 years ago

Agreed--that's a slippery slope. If we have non-Water, I think we should be good!
Done...

Also available in: Atom PDF