Bob,
After being out all day, I'm back at work. My intent was to begin a
systematic slow trek through all of the octahedral isomer types.
Unfortunately, I did not get far - SX4Y2. It's a strange problem, so I
thought that I better check in with you. By the way, because of this
problem early on, I went back to collecting stddev:
var i = {*}.atomIndex.max + 1;var t = {*}.atomIndex.min;select
within(branch, {atomIndex = i}, {atomIndex=t});var j =
{selected}.find('SMILES');var x = compare({selected}, {not selected},
@j, 'stddev');echo @x
Call this osrCompare below.
I loaded SF6; I duplicated it in the window to produce two identical
models of SF6: osrCompare ~0 ( 4.5E-8)
I made the trans Br isomer of both: osrCompare = NaN YIKES!
A quick look at the problem:
Original SMILES: [S@@]123456.F6.F1.F2.F3.F4.F5
Duplicated SMILES: [S@@]123456.F6.F1.F2.F3.*F5.F4*
Original SMILES (after Br's): [S@@]123456.F6.Br1.Br2.F3.F4.F5
Duplicated SMILES (after Br's): [S@@]123456.F6.Br1.Br2.F3.*F5.F4*
My duplication method is to do jmolGetPropertyAsString("extractModel",
"all") and then a load APPEND.
In the testing that I did, I never thought to test an obvious match. I
just decided that I should be thorough here.
Otis
Otis Rothenberger
chemagic.com
On 11/15/2010 10:19 AM, Otis Rothenberger wrote:
Bob,
For the time being, I'm simply trapping and replacing the responses
from your Jmol compare/isomer routine when coordination compounds are
loaded. I can automate this change over in JavaScript. The response
replacements allow identical/not identical/not same formula responses
for coordination compounds. This will allow me to do some quick checking.
I've played a bit, and this approach seems to work. Square planar is
fine. Octahedral will take more testing. I'll get back to you on this.
Otis
Otis Rothenberger
chemagic.com
On 11/15/2010 6:53 AM, Robert Hanson wrote:
Otis, I've also started the same thread at
<blueobelisk-smi...@lists.sourceforge.net
<mailto:blueobelisk-smi...@lists.sourceforge.net>>, and I'm hoping it
can be explained to me how you do this with T-shaped, see-saw, and
square pyramidal shapes. I think perhaps the answer is going to be
that you just leave out the missing groups and use explicit [...@tbn]
or [...@ohn] stereochemical descriptors (and possibly [...@thn] for
trigonal pyramidal). But I'm not sure about that, because, for example:
[...@oh2]1234.c1.br2.cl3.n4
gives the warning:
WARNING: Chiral center has wrong connectivity -- ignored
at the Daylight Depict site. Hope to hear about that soon.
On Mon, Nov 15, 2010 at 12:44 AM, Otis Rothenberger
<osrot...@chemagic.com <mailto:osrot...@chemagic.com>> wrote:
Bob,
I think I see it. If I'm analyzing @ or @@, group 1 is toward me.
Is that correct?
Yes, group 1 is toward you, and the rotation about the others is the
same as depicted in the @ symbol itself -- CCW for "@" . The first
and last groups must be axial. The ones in between are the ones that
are checked for rotation order.
Notice that the way this works for SMILES generation in Jmol is that
Jmol will now first list all "components" based on atoms that have
more than four attachments. For example, a particular isomer of
SF3IBrCl might be:
[...@]123456.i5.f1.br6.f4.f3.cl2
and a ferrocene model I have returns:
[fe]123456789%10...@h]1%11[c@h]...@h]4[c@H]3[C@@h]%112...@h]6%12[c@h]%1...@h]9[c@h]...@h]7%12
That is, we're creating a "generic" n-coordinate center and then just
listing the attached components. The nice thing about this is that
there's no significance to the order of groups in the individual
components -- it's just the connection numbers 1-10 that determines
the stereochemistry, and you can have as many of these connected
inorganic components as you need to get the job done.
The success here is based on not requiring any sort of "canonical" or
"unique" SMILES description for our purposes. Jmol can match
structures to stereochemical SMILES strings as well as stereochemical
SMILES strings to stereochemical SMILES strings without any need for
canonicalization. So that's all we need.
Except for your finding a bug possibly, Jmol should now be able to
match any stereochemical SMILES involving allenyl, tetrahedral,
trigonal bipyramidal, or octahedral stereochemistry descriptors, and
it will generate stereochemical SMILES strings for structures with
all of these shapes automatically.
------------------------------------------------------------------------------
Beautiful is writing same markup. Internet Explorer 9 supports
standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3.
Spend less time writing and rewriting code and more time creating great
experiences on the web. Be a part of the beta today
http://p.sf.net/sfu/msIE9-sfdev2dev
_______________________________________________
Jmol-users mailing list
Jmol-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/jmol-users
