{ "version": "https://jsonfeed.org/version/1.1", "title": "chem-bla-ics", "description": "Chemblaics (pronounced chem-bla-ics) is the science that uses open science and computers to solve problems in chemistry, biochemistry and related fields.", "home_page_url": "https://chem-bla-ics.linkedchemistry.info/", "feed_url": "https://chem-bla-ics.linkedchemistry.info/2006/12/30/modern-chemistry-in-cdk-beyond-two.json", "icon": "https://chem-bla-ics.linkedchemistry.info/assets/images/chem-bla-ics_logo.png", "language": "en", "authors": [ { "name": "Egon Willighagen", "url": "https://orcid.org/0000-0001-7542-0286", "_orcid": "0000-0001-7542-0286" } ], "items": [ { "id": "https://doi.org/10.59350/h2ytc-6pt51", "url": "https://chem-bla-ics.linkedchemistry.info/2006/12/30/modern-chemistry-in-cdk-beyond-two.html", "title": "Modern chemistry in the CDK: beyond the two-atom bond", "content_html": "

Rich recently blogged \nabout the limitations of the two-atom bond representation often used in chemoinformatics,\ntriggered by the four ferrocene entries in PubChem .\nIn reply to himself, Rich described FlexMol ,\nan XML language that can describe bond systems that involve more than two atoms.

\n\n

Obviously, the problems originates from the lack of mathematical knowledge of chemists: the\ncurrent chemoinformatics heavily depends on graph theory, where each atom is a vertex and each\nbond an edge. This has the advantage that we can borrow all algorithms that work with graph\nrepresentations, such as Dijkstra’s algorithm\nto find the shortest path between two vertices. Or, in chemical language, an algorithm to\ncalculate how many bonds two atoms are apart in a molecule.

\n\n

When discussing FlexMol, Rich mentions the work by Dietz (DOI:10.1021/ci00027a001),\nbut I would like to mention the PhD thesis of S. Bauerschmidt to this (see\nDOI:10.1021/ci9704423) done in Gasteiger’s group.\nDropping this ‘two-atom bond’ representation in favor of something that better describes compounds\nlike ferrocene, like the Dietz and Bauerschmidt approaches, has the unfortunate disadvantage of\nloosing compatibility with graph theory algorithms. Nevertheless, in order to take\nchemoinformatics to the next level, we have to address these issues. But hope is not lost, and\npeople are working on rewriting our toolkit of chemoinformatics algorithms to match such new\nrepresentations.

\n\n

CDK

\n\n

I will postpone analyzing the CDK for compatibility with such more modern\nrepresentations (look out for a CDK News article), and now just describe\nhow the CDK can be used for FlexMol/Dietz/Bauerschmidt representations. Consider\nthe four examples Rich gives \nin his blog. Here are the CDK ways of doing the same.

\n\n

For example, 1,3,5-cyclohexatriene:

\n\n
public IMolecule makeCycloHexaTriene() {\n  IMolecule cyclohexatriene = builder.newMolecule();\n\n  IAtom atomC0 = builder.newAtom(Elements.CARBON);\n    atomC0.setID(\"C0\"); atomC0.setHydrogenCount(1);\n  IAtom atomC1 = builder.newAtom(Elements.CARBON);\n    atomC1.setID(\"C1\"); atomC1.setHydrogenCount(1);\n  IAtom atomC2 = builder.newAtom(Elements.CARBON);\n    atomC2.setID(\"C2\"); atomC2.setHydrogenCount(1);\n  IAtom atomC3 = builder.newAtom(Elements.CARBON);\n    atomC3.setID(\"C3\"); atomC3.setHydrogenCount(1);\n  IAtom atomC4 = builder.newAtom(Elements.CARBON);\n    atomC4.setID(\"C4\"); atomC4.setHydrogenCount(1);\n  IAtom atomC5 = builder.newAtom(Elements.CARBON);\n    atomC5.setID(\"C5\"); atomC5.setHydrogenCount(1);\n\n  IBond bondB0 = builder.newBond(atomC0, atomC1, 1.0);\n    bondB0.setElectronCount(2);\n  IBond bondB1 = builder.newBond(atomC1, atomC2, 2.0);\n    bondB1.setElectronCount(4);\n  IBond bondB2 = builder.newBond(atomC2, atomC3, 1.0);\n    bondB2.setElectronCount(2);\n  IBond bondB3 = builder.newBond(atomC3, atomC4, 2.0);\n    bondB3.setElectronCount(4);\n  IBond bondB4 = builder.newBond(atomC4, atomC5, 1.0);\n    bondB4.setElectronCount(2);\n  IBond bondB5 = builder.newBond(atomC0, atomC5, 2.0);\n    bondB5.setElectronCount(4);\n\n  cyclohexatriene.addAtom(atomC0); cyclohexatriene.addAtom(atomC1);\n  cyclohexatriene.addAtom(atomC2); cyclohexatriene.addAtom(atomC3);\n  cyclohexatriene.addAtom(atomC4); cyclohexatriene.addAtom(atomC5);\n\n  cyclohexatriene.addBond(bondB0); cyclohexatriene.addBond(bondB1);\n  cyclohexatriene.addBond(bondB2); cyclohexatriene.addBond(bondB3);\n  cyclohexatriene.addBond(bondB4); cyclohexatriene.addBond(bondB5);\n\n  return cyclohexatriene;\n}\n
\n\n

Summarizing, the key thing is to use the IBond.setElectronCount() method.\nThe call is sort of redundant, as the CDK defaults to two electrons if not\nexplicitly given. This compound is, of course, benzene which we can represent\nlike this too:

\n\n
public IMolecule makeBenzene() {\n  IMolecule benzene = builder.newMolecule();\n\n  IAtom atomC0 = builder.newAtom(Elements.CARBON);\n    atomC0.setID(\"C0\"); atomC0.setHydrogenCount(1);\n  IAtom atomC1 = builder.newAtom(Elements.CARBON);\n    atomC1.setID(\"C1\"); atomC1.setHydrogenCount(1);\n  IAtom atomC2 = builder.newAtom(Elements.CARBON);\n    atomC2.setID(\"C2\"); atomC2.setHydrogenCount(1);\n  IAtom atomC3 = builder.newAtom(Elements.CARBON);\n    atomC3.setID(\"C3\"); atomC3.setHydrogenCount(1);\n  IAtom atomC4 = builder.newAtom(Elements.CARBON); \n    atomC4.setID(\"C4\"); atomC4.setHydrogenCount(1);\n  IAtom atomC5 = builder.newAtom(Elements.CARBON); \n    atomC5.setID(\"C5\"); atomC5.setHydrogenCount(1);\n\n  IBond bondB0 = builder.newBond(atomC0, atomC1);\n    bondB0.setElectronCount(2);\n  IBond bondB1 = builder.newBond(atomC1, atomC2);\n    bondB1.setElectronCount(2);\n  IBond bondB2 = builder.newBond(atomC2, atomC3);\n    bondB2.setElectronCount(2);\n  IBond bondB3 = builder.newBond(atomC3, atomC4);\n    bondB3.setElectronCount(2);\n  IBond bondB4 = builder.newBond(atomC4, atomC5);\n    bondB4.setElectronCount(2);\n  IBond bondB5 = builder.newBond(atomC0, atomC5);\n    bondB5.setElectronCount(2);\n\n  IBond bondingSystem = builder.newBond();\n    bondingSystem.setElectronCount(6);\n    bondingSystem.setAtoms(\n      new IAtom[] { atomC0, atomC1, atomC2, \n                    atomC3, atomC4, atomC5}\n    );\n\n  benzene.addAtom(atomC0); benzene.addAtom(atomC1);\n  benzene.addAtom(atomC2); benzene.addAtom(atomC3);\n  benzene.addAtom(atomC4); benzene.addAtom(atomC5);\n\n  benzene.addBond(bondB0); benzene.addBond(bondB1);\n  benzene.addBond(bondB2); benzene.addBond(bondB3);\n  benzene.addBond(bondB4); benzene.addBond(bondB5);\n  benzene.addBond(bondingSystem);\n\n  return benzene;\n}\n
\n\n

This version represents the delocalized aromatic pi-system as one IBond:\none with 6 electrons, and 6 associated atoms.

\n\n

The cyclopentadienyl anion is represented similarly:

\n\n
public IMolecule makeCycloPentadienylAnion() {\n  IMolecule cp = builder.newMolecule();\n\n  IAtom atomC0 = builder.newAtom(Elements.CARBON);\n atomC0.setID(\"C0\"); atomC0.setHydrogenCount(1);\n  IAtom atomC1 = builder.newAtom(Elements.CARBON);\n atomC1.setID(\"C1\"); atomC1.setHydrogenCount(1);\n  IAtom atomC2 = builder.newAtom(Elements.CARBON);\n atomC2.setID(\"C2\"); atomC2.setHydrogenCount(1);\n  IAtom atomC3 = builder.newAtom(Elements.CARBON);\n atomC3.setID(\"C3\"); atomC3.setHydrogenCount(1);\n  IAtom atomC4 = builder.newAtom(Elements.CARBON);\n atomC4.setID(\"C4\"); atomC4.setHydrogenCount(1);\n\n  IBond bondB0 = builder.newBond(atomC0, atomC1);\n    bondB0.setElectronCount(2);\n  IBond bondB1 = builder.newBond(atomC1, atomC2);\n    bondB1.setElectronCount(2);\n  IBond bondB2 = builder.newBond(atomC2, atomC3);\n    bondB2.setElectronCount(2);\n  IBond bondB3 = builder.newBond(atomC3, atomC4);\n    bondB3.setElectronCount(2);\n  IBond bondB4 = builder.newBond(atomC4, atomC0);\n    bondB4.setElectronCount(2);\n\n  IBond bondingSystem = builder.newBond();\n    bondingSystem.setElectronCount(6);\n  bondingSystem.setAtoms(\n    new IAtom[]{ atomC0, atomC1, atomC2, atomC3, atomC4}\n  );\n\n  cp.addAtom(atomC0); cp.addAtom(atomC1);\n  cp.addAtom(atomC2); cp.addAtom(atomC3);\n  cp.addAtom(atomC4);\n\n  cp.addBond(bondB0); cp.addBond(bondB1);\n  cp.addBond(bondB2); cp.addBond(bondB3);\n  cp.addBond(bondB4); cp.addBond(bondingSystem);\n\n  return cp;\n}\n
\n\n

And the final step in this series, is ferrocene:

\n\n
public IMolecule makeFerrocene() {\n  IMolecule ferrocene = builder.newMolecule();\n\n  IAtom atomC0 = builder.newAtom(Elements.CARBON);\n    atomC0.setID(\"C0\"); atomC0.setHydrogenCount(1);\n  IAtom atomC1 = builder.newAtom(Elements.CARBON);\n    atomC1.setID(\"C1\"); atomC1.setHydrogenCount(1);\n  IAtom atomC2 = builder.newAtom(Elements.CARBON);\n    atomC2.setID(\"C2\"); atomC2.setHydrogenCount(1);\n  IAtom atomC3 = builder.newAtom(Elements.CARBON);\n    atomC3.setID(\"C3\"); atomC3.setHydrogenCount(1);\n  IAtom atomC4 = builder.newAtom(Elements.CARBON);\n    atomC4.setID(\"C4\"); atomC4.setHydrogenCount(1);\n  IAtom atomC5 = builder.newAtom(Elements.CARBON);\n    atomC5.setID(\"C5\"); atomC5.setHydrogenCount(1);\n  IAtom atomC6 = builder.newAtom(Elements.CARBON);\n    atomC6.setID(\"C6\"); atomC6.setHydrogenCount(1);\n  IAtom atomC7 = builder.newAtom(Elements.CARBON);\n    atomC7.setID(\"C7\"); atomC7.setHydrogenCount(1);\n  IAtom atomC8 = builder.newAtom(Elements.CARBON);\n    atomC8.setID(\"C8\"); atomC8.setHydrogenCount(1);\n  IAtom atomC9 = builder.newAtom(Elements.CARBON);\n    atomC9.setID(\"C9\"); atomC9.setHydrogenCount(1);\n  IAtom iron = builder.newAtom(Elements.IRON);\n    iron.setID(\"Fe10\"); iron.setHydrogenCount(0);\n\n  IBond bondB0 = builder.newBond(atomC0, atomC1);\n    bondB0.setElectronCount(2);\n  IBond bondB1 = builder.newBond(atomC1, atomC2);\n    bondB1.setElectronCount(2);\n  IBond bondB2 = builder.newBond(atomC2, atomC3);\n    bondB2.setElectronCount(2);\n  IBond bondB3 = builder.newBond(atomC3, atomC4);\n    bondB3.setElectronCount(2);\n  IBond bondB4 = builder.newBond(atomC4, atomC0);\n    bondB4.setElectronCount(2);\n  IBond bondB5 = builder.newBond(atomC5, atomC6);\n    bondB5.setElectronCount(2);\n  IBond bondB6 = builder.newBond(atomC6, atomC7);\n    bondB6.setElectronCount(2);\n  IBond bondB7 = builder.newBond(atomC7, atomC8);\n    bondB7.setElectronCount(2);\n  IBond bondB8 = builder.newBond(atomC8, atomC9);\n    bondB8.setElectronCount(2);\n  IBond bondB9 = builder.newBond(atomC9, atomC5);\n    bondB9.setElectronCount(2);\n\n  IBond bondingSystem1 = builder.newBond();\n    bondingSystem1.setElectronCount(6);\n    bondingSystem1.setAtoms(\n      new IAtom[] {\n       atomC0, atomC1, atomC2, atomC3, atomC4, iron\n      }\n    );\n  IBond bondingSystem2 = builder.newBond(); \n    bondingSystem2.setElectronCount(6);\n    bondingSystem2.setAtoms(\n      new IAtom[] {\n        atomC5, atomC6, atomC7, atomC8, atomC9, iron\n      }\n    );\n  IBond bondingSystem3 = builder.newBond();\n    bondingSystem3.setElectronCount(6);\n    bondingSystem3.setAtoms(\n      new IAtom[]{\n        atomC0, atomC1, atomC2, atomC3, atomC4,\n        atomC5, atomC6, atomC7, atomC8, atomC9,\n        iron\n      }\n    );\n\n  ferrocene.addAtom(atomC0); ferrocene.addAtom(atomC1);\n  ferrocene.addAtom(atomC2); ferrocene.addAtom(atomC3);\n  ferrocene.addAtom(atomC4); ferrocene.addAtom(atomC5);\n  ferrocene.addAtom(atomC6); ferrocene.addAtom(atomC7);\n  ferrocene.addAtom(atomC8); ferrocene.addAtom(atomC9);\n  ferrocene.addAtom(iron);\n\n  ferrocene.addBond(bondB0); ferrocene.addBond(bondB1);\n  ferrocene.addBond(bondB2); ferrocene.addBond(bondB3);\n  ferrocene.addBond(bondB4);\n  ferrocene.addBond(bondB5); ferrocene.addBond(bondB6);\n  ferrocene.addBond(bondB7); ferrocene.addBond(bondB8);\n  ferrocene.addBond(bondB9);\n  ferrocene.addBond(bondingSystem1);\n  ferrocene.addBond(bondingSystem2);\n  ferrocene.addBond(bondingSystem3);\n\n  return ferrocene;\n}\n
\n\n

Now, you will note that this approach does not exactly follow Rich’s FlexMol examples: the\nskipped atom pair concepts in the FlexMol version of ferrocene. His example, more closely follows\nwhat we are likely to draw, while the CDK code above more closely follows the molecular orbital\nconcept. (I have to check to see how Dietz and Bauerschmidt did this.)

\n\n

As said, the real trick is to have the chemoinformatics toolkit that can work with this\nrepresentation, but I will save that for later. At least our algorithms to calculate the\nmolecular mass should work ;)

\n\n

References

\n
\n
Apodaca, R. (2006). A Molecular Language for Modern Chemistry: Getting Started with FlexMol. In Front Matter. Front Matter. https://doi.org/10.59350/s1vtx-e9q82 CrossRef Scholia
\n
Apodaca, R. (2006). Ferrocene and Beyond: A Solution to the Molecular Representation Problem. In Front Matter. Front Matter. https://doi.org/10.59350/pz3p6-fv247 CrossRef Scholia
\n
Apodaca, R. (2006). The Problem with Ferrocene. In Front Matter. Front Matter. https://doi.org/10.59350/xnt9b-80962 CrossRef Scholia
\n
Bauerschmidt, S., & Gasteiger, J. (1997). Overcoming the Limitations of a Connection Table Description:  A Universal Representation of Chemical Species. Journal of Chemical Information and Computer Sciences, 37(4), 705–714. https://doi.org/10.1021/ci9704423 CrossRef Scholia
\n
Dietz, A. (1995). Yet Another Representation of Molecular Structure. Journal of Chemical Information and Computer Sciences, 35(5), 787–802. https://doi.org/10.1021/ci00027a001 CrossRef Scholia
\n
", "summary": "Rich recently blogged about the limitations of the two-atom bond representation often used in chemoinformatics, triggered by the four ferrocene entries in PubChem . In reply to himself, Rich described FlexMol , an XML language that can describe bond systems that involve more than two atoms.", "date_published": "2006-12-30T00:00:00+00:00", "date_modified": "2025-02-16T00:00:00+00:00", "tags": ["cheminf","cdk"], "_references": [ { "url": "https://doi.org/10.1021/ci00027a001", "doi": "10.1021/ci00027a001" } , { "url": "https://doi.org/10.1021/ci9704423", "doi": "10.1021/ci9704423" } , { "url": "https://doi.org/10.59350/pz3p6-fv247", "doi": "10.59350/pz3p6-fv247" } , { "url": "https://doi.org/10.59350/xnt9b-80962", "doi": "10.59350/xnt9b-80962" } , { "url": "https://doi.org/10.59350/s1vtx-e9q82", "doi": "10.59350/s1vtx-e9q82" } ], "authors": [ { "name": "Egon Willighagen", "url": "https://orcid.org/0000-0001-7542-0286" } ] } ] }