Atom selection questions

Hi all, I have a question. We want to apply representations to specific water atoms. We currently select the atoms like this

atoms: (atomIds) => {
                return query.struct.generator.atomGroups({
                    'atom-test': query.core.set.has([
                        query.core.type.set(atomIds),
                        query.ammp('id')
                    ])
                });
            },

However this ends up slowing down our browser a lot (freezes the tab on my computer).

We compared it to this which ends up being much much faster even if we select the exact same atoms.

water: () => molstar.SelectionQueries.water.expression,

Is there any reason why index atom selections are so much slower than expressions? I would have assumed they would be faster since molstar would not have to search for what is water.

Also one more question. Is molstar 1-indexing atoms? I initially selected atoms with 0-based index but it always missed the last atom, then I switched to 1-based indexing, but then we found some sdf file where the 1-based index doesn’t work either.

[dsehnal]

The performance issue is fixed in 2.2.2.

Using sourceIndex() vs ammp('id'):

• ammp('id') (which is shorthand for struct.atomProperty.macromolecular.id()) is the value of the _atom_site.id column in mmCIF (the 2nd column in the PDB file). This is usually 1-based index of the atom in the file)
• struct.atomProperty.core.sourceIndex() is zero-based index of the ATOM/HETATM record as it appears in the input file.

[stefdoerr]

Indexing works like a charm! Will need to test now 2.2.2. Thanks a lot!