The IUPAC Chemical Identifier

Transcription

1 The IUPAC Chemical Identifier Steve Stein, Steve eller, Dmitrii Tchekhovskoi ational Institute of Standards and Technology Gaithersburg, MD, USA CAS/IUPAC Conference on Chemical Identifiers and XML for Chemistry Columbus, July 1, 2002

2 IUPAC & Chemical Identity Mission International, open standards for chemical communication Printed Media omenclature uman communication ules for structure to name conversion Digital Media Identifier Computer communication ules for structure to identifier conversion Freed from restrictions of pronouncibility Freed from ring index

3 Chemical Identifiers Structures Connection Tables Trivial ames Systematic ames Index umbers

4 Too Many Identifiers Structure diagrams various conventions contain too much information Connection Tables MolFiles, Smiles, SDAL,.. Pronounceable names IUPAC, CAS, trivial Index umbers EIECS, FEMA, DT, TECS, CAS, Beilstein, USP, TECS, EEC, CA, CI, U, USAF

5 What kind of Identifier is needed? Exactly one Identifier per structure Defined by algorithms Comprehensive penly available Implemented

6 equirements Different compounds have different identifiers All distinguishing structural information is included = IChI - 1 = IChI - 2

7 equirements ne compound has only one identifier o unnecessary information is included + = = = Same IChI

8 IChI Scope First Version Discrete, covalently bonded compounds foundation for other classes Isotopes Stereochemistry sp3 - tetrahedral Z/E - double bond Tautomers

9 3 Steps to IChI ormalize Input Structure Implement chemical rules Canonicalize (label the atoms) Equivalent atoms get the same label Serialize the Labeled Structure A unique series of bytes

10 MALIZATI

11 Simplifications Ignore Electron Density Double/triple bonds, dd-electrons, Charges Still use for Z/E stereo perception Free otation Around Single Bonds Divide IChI into Layers

12 Ignore Electron Density ot required for compound identification Distinguishes excited states Avoids problems Delocalization, aromaticity, zwitterions,

13 conventional C 2 C C C 2 redundant C C C 2 IChI 2 C

14 +

15 C C - + C - C - + C C C Münchnones

16 Assume Free otation Around Single Bonds = X X = X X Ignore Conformation

17 LAYES

18 Divide into Layers Separate ame into Fragments by Connectivity Isotopes Stereochemistry Tautomerism

19 Basic Layer Simple Connectivity Just atoms and their neighbors Ignore everything else obust basic identifier

20 Isotopes Treat isotopes as distinct atom types D = D D D

21 Stereochemistry Double Bond (Z/E) from coordinates or bonding Tetrahedral (sp 3 ) in/out bonds or x,y,z coordinates

22 Varieties of Double Bond Isomers

23 sp 3 (tetrahedral) stereoisomers C 3 C 2 F F F F F F Stereodescriptor needed

24 Identify Stereogenic Centers C * Speed up processing elpful for chemists

25 Basic Tautomer Layer -migration between 1,3 heteroatoms [

26 Tautomers Q L Q L L, =,, S, Se, Te Q = C,, S, P, also ' ' '' ''

27 Electronic Layer Simply Store et Charge Possiblities: eutral -1 (anion) +1 (radical cation) +2 (doubly charged) Electronic State?

28 UTPUT

29 IChI utput 9 possible fields Basic ## Isotopic ## Stereo ## Stereo ## Tautomeric ## Isotopic ## Stereo ## Stereo ## Electronic ##

30 Possible utput Format Example: Benzene epresent atoms as sequence number in formula C66 = C C C C C C tags Basic Layer: <basic>c </basic>

31 ther utput Information nly For user verification Label true stereogenic atoms Identify equivalent atoms Warnings Unusual valences Unrecognized input eversibility Information Coordinates Electron density Positions of double/triple bonds, charges, odd electrons

32 TESTIG

33 Mathon,. Sample Graphs for Isomorphism Testing Congressus umerantium V21, pp , 1978

34

35 PBLEMS

36 Two Fundamental Problems Chemists Different ways to represent the same thing Different definitions of tautomerism Different guesses Chemicals Structures can depend on conditions Tautomers can depend on conditions

37 When to allow double bond stereoisomerism? free restricted Proposed: If a bond can be single, no Z/E stereo allowed Considered: Allow users to override default behavior

38 Drawing Standard eeded? Bond/o bond Cl Cl Cl + C a C a C a (0) + C 3 C 3 C 3 P P + B B

39 Allow Full eversibility? Coordinates Structure display riginal bonds and charges For display and future use riginal numbering Map to input data

40 IChI What can t it do? Discover that two structures with different connectivity represent the same compound Unless they are tautomers Predict potential for Z/E isomerism in open shell conjugated networks Cannot predict rotational barriers Fix improperly entered data Guarantees wrong IChI for bad data Properly treat non-covalent bonding Coordinate bonds epresent exotic stereochemistry

41 Version I Implement All ormalization ules 12/02 Test against available data sets 3/03 Final External Testing and efinement 7/03 Documentation, source, executable 12/03? pen discussions ichi-l@list.rsc.org

42 Future Extensions rganometallics Coordinate bonds ther Stereo Forms on-atom centered Conformations ydrogen Bonding Polymers/Macromolecules Compound Classes Markush structures