>
> With regards to an "Open" CIP -- I strongly suggest not going there. If
> you are seriously interested in this, join/form an IUPAC project.
For me the main motivation is to not reinvent the wheel, or perhaps not
reinvent the wheel worse than it already was. The less people have to think
and worry about CIP the better.
Maybe an IUPAC project is the way to go there... but a stepping stone is to
have a 'standard/verified' implementation (i.e. Open CIP) first. You've
done well on your implementation and it may be one to recomend as a
'verified' version but if we do nothing in 2 years time someone else will
come along and waste more time with the same bugs/corner cases.
- You've still not done the iteration quite right, you only give 4 labels
for this one: [C@]3([C@]1([H])CC[C@@]([H])(C)CC1)([C@]2([H])CC[C@
@]([H])(C)CC2)CC[C@]([H])(C)CC3
- What CIP labels do you get for CHEBI:51439
<http://www.ebi.ac.uk/chebi/searchId.do?chebiId=CHEBI%3A51439>? I believe a
correct implementation should never finish... or at least run of memory
when building the digraph.
On 9 May 2017 at 19:48, Robert Hanson <hans...@stolaf.edu> wrote:
> OK, 967 lines. This now includes chiral bridgehead nitrogens and is just
> missing a small piece of code for integrating M/P and seqCis/seqTrans into
> Rule 4b. The algorithm is solid. No particular issues other than that. Does
> not implement atom-number averaging for mancude rings, but does remove
> Kekule dependencies for aromatic rings. This last bit required a slight
> additional "friendly amendment" to Rule 1b, which I don't think was crafted
> with aromatics in mind. Pseudocode is still pretty much the same as
> initially described. There are no issues with multi-center dependencies.
>
> Validates on the full 236-compound Chapter 9 validation suite from
> ACD/Labs except:
>
> var skip = ({27 229}) || // ignore -- BB has E/Z only; missing chirality
> ({95 96 98 99 100 101 102 103 104 108 109 110 111 112 200}) ||
> // trigonal planar, square planar, or hypervalent
> ({32 33}) || // helicene
> ({212 213})|| // chiral conformation 1,4-benzene in a ring
> ({38}) || // ignoring -- Mancude for cyclopentadienyl -- will
> require some thought
> ({170}) // failing (mixed Rule 4b)
>
> In the process, I have found four erroneous assignments in IUPAC Blue Book
> 2013. These are being checked by IUPAC. There's actually quite a large
> errata page for that book already.
>
> With regards to an "Open" CIP -- I strongly suggest not going there. If
> you are seriously interested in this, join/form an IUPAC project.
>
> Bob
>
>
> On Sat, Apr 29, 2017 at 2:54 PM, Robert Hanson <hans...@stolaf.edu> wrote:
>
>> OK, final update for a while. (816 lines, Wolf. I have no idea how that
>> compares to yours or others' algorithms.)
>>
>> Open-source, validated IUPAC 2013 preferred IUPAC name (PIN)
>> stereochemical designations.
>>
>> Jmol.___JmolVersion="14.15.2" // 4/29/17
>>
>> -- 816 lines
>> -- validation data are at https://sourceforge.net/p/jmol
>> /code/HEAD/tree/trunk/Jmol-datafiles/cip/
>> -- validates for 160 structures (some duplicates; both cip_examples.zip
>> and stereo_test_cases.sdf)
>> -- validates for all cases considered:
>> -- simple R/S and E/Z
>> -- small-ring removal of E/Z
>> -- parallel-strand Rule 4b and Rule 5 (Mata)
>> -- pseudochiral r/s and m/p
>> -- odd and even cumulenes
>> -- atropisomers
>> -- P, S, As, Se, Sb, Te, Bi, Po trigonal pyramidal and tetrahedral
>> -- imine and diazine E/Z chirality
>>
>> The algorithm will fail for some more complex nested aspects of Rule 4b.
>> I decided to be satisfied for now with only those examples in IUPAC Blue
>> Book 2013 Chapter 9. My understanding is that even ACD/Labs did not fully
>> implement Rules 4 and 5 much beyond that.
>>
>> Working version in JavaScript can be tested at
>> https://chemapps.stolaf.edu/jmol/jsmol/jsmetest2.htm
>> Binary and source at https://sourceforge.net/projects/jmol/files/Jmol/
>>
>> A great challenge for April!
>>
>> Bob
>>
>>
>>
>> On Thu, Apr 27, 2017 at 12:21 AM, Robert Hanson <hans...@stolaf.edu>
>> wrote:
>>
>>> Rule 5 is done. Fully validating using the first validation set that
>>> Mikko sent me (86 compounds, roughly, some 2D/3D duplicates). I'm sure
>>> there are more cases it needs testing with, though.
>>>
>>> My algorithm implementation handles Rules 4 and 5 lexicographically so
>>> that a simple Array.sort(String[]) does the job. Kind of interesting,
>>> perhaps.
>>>
>>> 763 lines, Wolf.
>>>
>>> Bob
>>>
>>>
>>>
>>
>>
>> --
>> Robert M. Hanson
>> Larson-Anderson Professor of Chemistry
>> St. Olaf College
>> Northfield, MN
>> http://www.stolaf.edu/people/hansonr
>>
>>
>> If nature does not answer first what we want,
>> it is better to take what answer we get.
>>
>> -- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
>>
>>
>
>
> --
> Robert M. Hanson
> Larson-Anderson Professor of Chemistry
> St. Olaf College
> Northfield, MN
> http://www.stolaf.edu/people/hansonr
>
>
> If nature does not answer first what we want,
> it is better to take what answer we get.
>
> -- Josiah Willard Gibbs, Lecture XXX, Monday, February 5, 1900
>
>
> ------------------------------------------------------------
> ------------------
> Check out the vibrant tech community on one of the world's most
> engaging tech sites, Slashdot.org! http://sdm.link/slashdot
> _______________________________________________
> Blueobelisk-discuss mailing list
> Blueobelisk-discuss@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/blueobelisk-discuss
>
>
------------------------------------------------------------------------------
Check out the vibrant tech community on one of the world's most
engaging tech sites, Slashdot.org! http://sdm.link/slashdot
_______________________________________________
Blueobelisk-discuss mailing list
Blueobelisk-discuss@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/blueobelisk-discuss
