From: Andrew Dalke (dalke_at_dalkescientific.com)
Date: Fri Jan 07 2005 - 17:36:50 CST

John wrote:
> I could definitely see adding a new selection keyword to allow
> selection
> by the index values from the original file, though the responsbility
> of watching out for the items I list above would then be on the
> shoulders
> of the user. I don't have time to mess with this for VMD 1.8.3, but
> that
> might be worth doing for a subsequent revision.

One thing to point out is the PDB spec regards that as a serial number
field where the first record is 1 and the numbers proceed sequentially.

As such the difference between "serial" (the hypothetical name
used for this field) and "index" (the 0-based version of that field)
is always 1.

It's possible for someone to use a different numbering scheme for
those fields but as John pointed out those numbers are best ignored.
I don't think there should be a field to access those number exactly
as given. (Should "00001" be considered the same as " 1"?)

So I think the solution to pl's problem could be to just add a
new integer function to the atom selection code to handle "serial".

                                        Andrew
                                        dalke_at_dalkescientific.com