The internet has provided access to unprecedented quantities of data. In the domain of chemistry specifically over the past decade the web has become populated with tens of millions of chemical structures and related properties, both experimental and predicted, together with tens of thousands of spectra and syntheses. The data have, to a large extent, remained disparate and disconnected. In recent years with the wave of Web 2.0 participation any chemist can contribute to both the sharing and validation of chemistry-related data whether it be via Wikipedia, the online encyclopedia, or one of the multiple public compound databases. Toxicologists commonly wish to source data, either for reference purposes, to support the development of models or, when experimental data are not available, predicted data will suffice. This presentation will offer a perspective of the type and quality of chemistry data available today, our experiences of building the ChemSpider public compound database to link together chemistry on the internet and our efforts to both encourage and enable even greater integration and connectivity for chemistry data for the community.

1. ChemSpider as an Integration
Hub for Interlinked
Chemistry Data
Antony Williams
SETAC
November 18th 2013

2. How Much Data Online?
• How much data regarding environmental
toxicology and chemistry is online?
• How can it all be mapped together?

3. A Grand Challenge….
• Let’s map together all historical chemistry
data and build systems to integrate new data
• Let’s integrate chemistry, toxicology and
biology data and add in disease data too
• Lets model the data and see if we can
extract new relationships – quantitative and
qualitative
• Let’s make it all available on the web

5. What about this….
• We’re going to map the world
• We’re going to take photos of as many
places as we can and link them together
• We’ll let people annotate and curate the map
• Then let’s make it available free on the web
• We’ll make it available for decision making
• Put it on Mobile Devices, Give it Away

6. The World of Online Chemistry
•
•
•
•
•
•
•
•
Property databases
Compound aggregators
Screening assay results
Scientific publications
Encyclopedic articles (Wikipedia)
Metabolic pathway databases
ADME/Tox data – eTOX for example
Blogs/Wikis and Open Notebook Science

7. How to Map Data Together
• Download the structure representations
and map together at the structure level
• Integrate and mesh chemical names,
chemical properties, analytical data
• Carry URL links and retain external links to
original data sets (assume no link decay)
• It sounds easy….

8. ChemSpider
• Build a HUB connecting as many data
sources as possible
• NOT to harvest all data from each data source
• Today we have >29 million unique chemicals
from >500 data sources
• Focus on improving data quality
• Allow users to enhance, curate and annotate

9. RSC’s ChemSpider

10. Identifiers are very useful! But
what when they are “closed”

11. CAS Numbers Validation?

12. Various Registration Numbers

13. Mappings and Inconsistencies
Imatinib
Mesylate
ChemSpider
Drugbank
PubChem

14. The InChI Identifier

15. InChIStrings Hash to InChIKeys

16. Vancomycin – Search the
Internet

17. Vancomycin
Search Molecular
SKELETON
Search Full Molecule

18. Full Skeleton Search: 529 Hits

19. Full Molecule Search: 294 Hits

20. Historical Data for reference
• As evidence that InChI is proliferating and
data is improving:
• Three years ago there were only 104 hits
on the complete InChI online
• Only 4 were correct

21. What you might not know about
Chemistry Databases on the Internet

22. NCGC Pharma Collection

23. NCGC Pharma Collection

24. NCGC Pharma Collection

25. PHYSPROP Database
• The freely downloadable
database under the EPI
Suite prediction software
• Very Basic filters suggest
data quality issues

26. The Stereochemistry challenge.
12500 chemicals with “missed” stereo

27. NIST Webbook

28. PubChem

29. Patents

30. Patents

31. But Chemspider is curated right?

32. Originally 15 compounds “called” Yohimbine
54 Skeletons for Yohimbine

33. Crowdsourced Curation
• Crowd-sourced curation: identify/tag errors,
edit names, synonyms, identify records to
deprecate

34. Search “Vitamin H”

35. “Curate” Identifiers

36. “Curate” Identifiers

37. “Curate” Identifiers

38. Chemical name dictionaries for:
• Text-mining (publications, patents)
• Used to index PubMed and link Google Patents
• Linking to other databases – think Biology!
• When structures are not available names link
• Searching the web
• Names link to structures link to InChIs

39. I want to know about “Vincristine”

40. Vincristine: Identifiers to link

41. Vincristine: Vendors and Sources
Linked by Structure

42. Vincristine: Patents
Linked by Name

43. Vincristine: Articles
Linked by Name

44. What needs to happen?
• Standards
• Standardization of structures
• More sharing of data – downloadable data
collections for mapping, meshing and integration
• InChI adoption
• Collaboration
• Stop reinventing the wheel
• Share data, share efforts and speed the
process

45. Adopting Modified FDA Rules

46. Nitro groups

47. Salt and Ionic Bonds

48. Ammonium salts

49. What if we could capture it all?
Digitally Enhancing the RSC Archive

50. Start with data in publications

51. Text Mining
The N-(β-hydroxyethyl)-N-methyl-N'-(2-trifluoromethyl-1,3,4thiadiazol-5-yl)urea prepared in Example 6 , thionyl chloride
( 5 ml ) and benzene ( 50 ml ) were charged into a glass
reaction vessel equipped with a mechanical stirrer,
thermometer and reflux condenser.
The reaction mixture was heated at reflux with stirring , for a
period of about one-half hour.
After this time the benzene and unreacted thionyl chloride
were stripped from the reaction mixture under reduced
pressure to yield the desired product N-(β-chloroethyl)-Nmethyl-N'-(2-trifluoromethyl-1,3,4-thiaidazol-5-yl)urea as a
solid residue

52. ChemSpider Reactions

53. Turn “Figures” Into Data
FIGURE
EXTRACTED DATA

54. Conclusions
• There are some amazing online resources for
environmental toxicology and chemistry already!
• ChemSpider has an important role in quality
data and linking resources
• Crowdsourced deposition, validation and
curation works
• Standards are an important part of data linking
• MORE collaboration and data sharing can
benefit us all

55. Thank you
Email: williamsa@rsc.org
Twitter: @ChemConnector
Personal Blog: www.chemconnector.com
SLIDES: www.slideshare.net/AntonyWilliams