![\"\"](\"/assets/images/bjoc_bioschemas.png\")
Two weeks ago, the Beilstein Institute announced Bioschemas support in their journals:
\n\n\n\n\nWe streamline the discoverability of your research by incorporating machine-readable chemical information into many of our published articles.\nThis includes the conversion of chemical structures from submitted ChemDraw files to InChI strings and validating them using open-source tools.
\n
The idea is far from new and has been around for two decades. But the two Beilstein journals\n(both diamond Open Access), actually integrated into their active publishing model.\nThat has been trialed and put in action before. For example, there was (is?) Project Prospect\n(2007), chemical structure annotation in Nature Chemistry\n(2009), SMILES in the ACS Journal of Medicinal Chemistry\n(2014) (doi:10.1021/jm5002056),\nand FAIR chemical structures in the Journal of Cheminformatics (2021) (doi:10.1186/s13321-021-00520-4).
\n\nBut this announcement is a new step. I like how validation of the chemical structures is part of the approach, and I like\nhow they use the Bioschemas extention of schema.org. The last because\nthey use two Bioschemas types/profiles that contributed to or initiated, respectively: MolecularEntity\nand ChemicalSubstance.
\n\nFirst stop for me is to check the schema.org annotation with a validation tool, like Google’s Rich Results Test.\nThat gives an idea how they may have have their search engine pick it up. The test article I was given on LinkedIn is\nXiao et al.’s Molecular diversity of the reactions of MBH carbonates of isatins and various nucleophiles\n(doi:10.3762/bjoc.21.21) in the Beilstein Journal of Organic Chemistry,\nand we indeed see the schema.org annotation show up:
\n\n
And because of the use of open standards, extracting the information is not so hard with, for example here,\nBacting (doi:10.21105/joss.02558), based on a 2022 script from the NanoSafety Cluster\nprojects NanoCommons and SbD4Nano:
\n\n@Grab(group='io.github.egonw.bacting', module='managers-rdf', version='1.0.4')\n@Grab(group='io.github.egonw.bacting', module='managers-ui', version='1.0.4')\n@Grab(group='io.github.egonw.bacting', module='net.bioclipse.managers.jsoup', version='1.0.4')\n\nbioclipse = new net.bioclipse.managers.BioclipseManager(\".\");\nrdf = new net.bioclipse.managers.RDFManager(\".\");\njsoup = new net.bioclipse.managers.JSoupManager(\".\");\n\narticles = [\n args[0]\n]\n\nkg = rdf.createInMemoryStore()\n\nfor (article in articles) {\n htmlContent = bioclipse.download(article)\n\n htmlDom = jsoup.parseString(htmlContent)\n\n // application/ld+json\n\n bioschemasSections = jsoup.select(htmlDom, \"script[type='application/ld+json']\");\n\n for (section in bioschemasSections) {\n bioschemasJSON = section.html()\n rdf.importFromString(kg, bioschemasJSON, \"JSON-LD\")\n }\n}\n\nturtle = rdf.asTurtle(kg);\n\nprintln \"#\" + rdf.size(kg) + \" triples detected in the JSON-LD\"\n// println turtle\n\n\nsparql = \"\"\"\nPREFIX schema: <http://schema.org/>\nSELECT ?entity ?inchikey ?smiles WHERE {\n ?entity a schema:MolecularEntity .\n OPTIONAL { ?entity schema:inChIKey ?inchikey }\n OPTIONAL { ?entity schema:smiles ?smiles }\n}\n\"\"\"\n\nresults = rdf.sparql(kg, sparql)\n\nfor (i=1;i<=results.rowCount;i++) {\n println \"${results.get(i, \"inchikey\")}\\t${results.get(i, \"smiles\")}\"\n}\nThe output is a simple table:
\n\nMGAPJMNPGGTFHJ-JEIPZWNWSA-N CN1C(=O)/C(=C/2\\C3=CC(=CC=C3N(CC4=CC=CC=C4)C2=O)Cl)/C(=P(C5=CC=CC=C5)(C6=CC=CC=C6)C7=CC=CC=C7)C1=O\nXEWMQVUVGAHESA-UHFFFAOYSA-N CC1=CC=C(C=C1)NC2=C(C3C4=CC(=CC=C4N(CC5=CC=CC=C5)C3=O)C)C(=O)N(C)C2=O\nUVTJORFYHPGJDZ-PYCFMQQDSA-N CCCCN1C2=CC=C(C)C=C2/C(=C(\\C#N)/CNC3=CC=C(C)C=C3)/C1=O\nILWGDUYVQRAMMG-PGMHBOJBSA-N CCCCN1C2=CC=C(C)C=C2/C(=C(\\C#N)/CNC3=CC=C(C=C3)Cl)/C1=O\nCAFIBKBZWJFZCW-FXBPSFAMSA-N CCCCN1C2=CC=C(C)C=C2/C(=C(\\C#N)/CNC3=CC=CC=C3)/C1=O\nUOJSFLANMVIMBV-UHFFFAOYSA-N CCCCN1C2=CC=C(C)C=C2C(C3=C(C(=O)N(C)C3=O)NC4=CC=C(C=C4)Cl)C1=O\nVNJBTGZXAGHCSO-OAPYJULQSA-N COC(=O)/C(=C\\1/C2=C(C=CC=C2)N(CC3=CC=CC=C3)C1=O)/C=P(C4=CC=CC=C4)(C5=CC=CC=C5)C6=CC=CC=C6\nKJXQRAKSOANQTJ-GFMRDNFCSA-N CC1=CC=C(C=C1)NC/C(=C\\2/C3=C(C=CC=C3)N(CC4=CC=CC=C4)C2=O)/C#N\nIGEBJMZDOPBFGF-UHFFFAOYSA-N CCCCN1C2=CC=C(C)C=C2C(C3=C(C(=O)N(C)C3=O)NC4=CC=CC=C4)C1=O\nSSANVPNESOMKOM-AWQADKOQSA-N C1=CC=C(C=C1)CN2C3=CC=C(C=C3/C(=C(/C#N)\\C=P(C4=CC=CC=C4)(C5=CC=CC=C5)C6=CC=CC=C6)/C2=O)Cl\nGEHWHSHQSIOZKL-NVQSTNCTSA-N CCCCN1C2=CC=C(C=C2/C(=C\\3/C(=P(C4=CC=CC=C4)(C5=CC=CC=C5)C6=CC=CC=C6)C(=O)N(C)C3=O)/C1=O)Cl\nPALRSQOHFLRWDH-UHFFFAOYSA-N CCCCN1C2=CC=C(C)C=C2C(C3=C(C(=O)N(C)C3=O)NC4=CC=C(C=C4)OC)C1=O\nKBFODZMDSAFLFR-UHFFFAOYSA-N CN1C(=O)C(=C(C1=O)NC2=CC(=CC=C2)Cl)C3C4=CC(=CC=C4N(CC5=CC=CC=C5)C3=O)Cl\nJCGAVVZYXDJPBU-GFMRDNFCSA-N CC1=C(C=CC=C1)NC/C(=C\\2/C3=C(C=CC=C3)N(CC4=CC=CC=C4)C2=O)/C#N\nDZFPCPDEQGLPLY-UHFFFAOYSA-N CCCCN1C2=CC=C(C)C=C2C(C3=C(C(=O)N(C)C3=O)NC4=CC=C(C)C=C4)C1=O\nXMRNJCJUOXYXJU-DAFNUICNSA-N CC1=CC=C(C=C1)NC/C(=C\\2/C3=CC(=CC=C3N(CC4=CC=CC=C4)C2=O)C)/C#N\nSSDSNBBHEUUKGI-UHFFFAOYSA-N CC1=CC=C2C(=C1)C(C3=C(C(=O)N(C)C3=O)N(C)C4=CC=CC=C4)C(=O)N2CC5=CC=CC=C5\nUSFYPRDMNXMWPO-UHFFFAOYSA-N CCCCN1C2=CC=C(C)C=C2C(C3=C(C(=O)N(C)C3=O)NC4=CC=C(C=C4)Br)C1=O\nXYHTWFULRHTEAG-MUGXBBEHSA-N CCCCN1C2=CC=C(C)C=C2/C(=C(/C#N)\\C=P(C3=CC=CC=C3)(C4=CC=CC=C4)C5=CC=CC=C5)/C1=O\nXALDZIBHNNIVAM-UHFFFAOYSA-N CCCCN1C2=CC=C(C)C=C2C(C3=C(C(=O)N(C)C3=O)NC4=C(C=CC=C4)O)C1=O\nTUTWQHBRQPMLME-OAPYJULQSA-N COC(=O)/C(=C\\1/C2=CC(=CC=C2N(CC3=CC=CC=C3)C1=O)Cl)/C=P(C4=CC=CC=C4)(C5=CC=CC=C5)C6=CC=CC=C6\nIYEHFTMZZMIPRU-UHFFFAOYSA-N CC1=CC=C(C=C1)NC2=C(C3C4=CC(=CC=C4N(CC5=CC=CC=C5)C3=O)Cl)C(=O)N(C)C2=O\nKBSDGNPLIPXCEX-UHFFFAOYSA-N CCCCN1C2=CC=C(C)C=C2C(C3=C(C(=O)N(C)C3=O)NCC4=CC=CC=C4)C1=O\nBQGIUMITIGHBSD-UHFFFAOYSA-N CCCCNC1=C(C2C3=CC(=CC=C3N(CC4=CC=CC=C4)C2=O)C)C(=O)N(C)C1=O\nPNSOLOPHIVUPOZ-MNDPQUGUSA-N CCCCNC/C(=C\\1/C2=CC(=CC=C2N(CCCC)C1=O)C)/C#N\nHLTBKJRJOIZCMJ-PYCFMQQDSA-N CCCCN1C2=CC=C(C)C=C2/C(=C(\\C#N)/CN(C)C3=CC=CC=C3)/C1=O\nFFLHFLUBMRBQTB-UHFFFAOYSA-N CCCCN1C2=CC=C(C=C2C(C3=C(C(=O)N(C)C3=O)NC4=CC=C(C)C=C4)C1=O)F\nFOQOVOLYYARWPA-NKFKGCMQSA-N C1=CC=C(C=C1)CN2C3=C(C=CC=C3)/C(=C(\\C#N)/CNC4=CC(=CC=C4)Cl)/C2=O\nKLEPCAQFOXJLNV-UHFFFAOYSA-N CC1=C(C=CC=C1)NC2=C(C3C4=CC(=CC=C4N(CC5=CC=CC=C5)C3=O)Cl)C(=O)N(C)C2=O\nThat also made me realize that there are not chemical names in the annotation. That would be really useful to move things\nforward. Then again, PubChem will likely just generate the IUPAC name, since they have access to such software anyway.\nThey have teamed up with PubChem which will index it, but I will be interested in seeing how to use this for\nmain subject annotation in Wikidata.
A final note for now, the model they use is annotate the article with chemical substances (ChemicalSubstance) with\n(one or more?) molecular entities (`MolecularEntity’). That is a model that scales well to their other journal,\nthe Beilstein Journal of Nanotechnology. But scraping that is for another post.