Docs improvement on WaterBridge

[cbouysset]

As mentioned by @yuyuan871111

• waters should have CONECT record in the topology file, or bonds guessed

Edit: removed

• segments are duplicated, maybe have a way to automatically update segindices to deduplicate them?

[yuyuan871111]

• segments are duplicated, maybe have a way to automatically update segindices to deduplicate them?

To clarify this, due to the limitation of 4 digits for resid in the PDB format file, two different water molecules might have the same resid in the same chain/segment. (see the example below)

[cbouysset]

for that example, what's the output of

wat_1917 = u.select_atoms("resname TIP3 and resnum 1917")
print(
    wat_1917.residues,
    set(wat_1917.chainIDs),
    set(wat_1917.residues.segindices),
)

[yuyuan871111]

for that example, what's the output of

wat_1917 = u.select_atoms("resname TIP3 and resnum 1917")
print(
wat_1917.residues,
set(wat_1917.chainIDs),
set(wat_1917.residues.segindices),
)

The output is <ResidueGroup [<Residue TIP3, 1917>]> {''} {2}.

[cbouysset]

Sorry I meant

print(len(wat_1917.residues))

So even with the segment identifier the waters are not unique... That will need fixing

[yuyuan871111]

I do more tests and check with the coordinates. I found MDAnalysis will automatically add 10000.

[cbouysset]

And does it do the same with resid instead of resnum?

[yuyuan871111]

And does it do the same with resid instead of resnum?

Yes

[cbouysset]

Ok so we should be all good for this duplicated ResidueId issue then?

[yuyuan871111]

Ok so we should be all good for this duplicated ResidueId issue then?

Yes, I think so.