1. Open PHACTS “Data integration for all” Andrew Leach

2. Task, workflow and results AUREUS search targets: voltage-gated potassium channels Apply filters (MW, cLogP, Lipinski + remove undesirable target) ⇒ ~1000 molecules Similarity searches (RG, TP, Daylight) Cluster analysis ⇒ ~ 10000 molecules selected IonWorks © single shot screening 240 single shot hits progressed into full curve assay 5 full curve actives (in at least one test occasion) Series for lead optimisation Stefan Senger, ca. 2004 Task: create a focussed set to identify leads against voltage- gated potassium channels

3. We (may) know where the data is, but integrating is a pain, bespoke, and often only for experts Q: Identify all oxidoreductase inhibitors with an activity <100nM in both mouse and human Q: The current Factor Xa lead series is characterised by substructure X. Retrieve all bioactivity data in serine protease assays for molecules that contain substructure X. Q: For a given interaction profile, give me compounds similar to it. ChEMBL DrugBank Gene Ontology Wikipathways Uniprot ChemSpider UMLS ConceptWiki ChEBI etc. Internal

4. The Innovative Medicines Initiative Biggest public-private partnership in area of medicine Collaboration between European Commission and European Federation of Pharmaceutical Industries and Associations (EFPIA) Promotion of medical innovation in Europe Tackle key bottlenecks Recognises “in kind” contributions Focus on key problems –Efficacy, Safety, Education & Training, Knowledge Management

5. Public Domain Drug Discovery Data Pharma are accessing, processing, storing & re-processing Why repeat at each company? GSK AZ Pfizer Merck

6. Information Tombs –Built for primary use-case –Tailored indexes –Tailored GUIs –Unique language & metadata –Poor interoperability/integration LiteratureHRSynthesisPortfolioSARDocsSafetyIn vivoEtc

7. Pfizer Limited – Coordinator Universität Wien – Managing entity Technical University of Denmark University of Hamburg, Center for Bioinformatics BioSolveIT GmBH Consorci Mar Parc de Salut de Barcelona Leiden University Medical Centre Royal Society of Chemistry Vrije Universiteit Amsterdam Spanish National Cancer Research Centre University of Manchester Maastricht University Aqnowledge University of Santiago de Compostela Rheinische Friedrich-Wilhelms-Universität Bonn AstraZeneca GlaxoSmithKline Esteve Novartis Merck Serono H. Lundbeck A/S Eli Lilly Netherlands Bioinformatics Centre Swiss Institute of Bioinformatics ConnectedDiscovery EMBL-European Bioinformatics Institute Janssen OpenLink Project Partners

8. A use-case driven approach, focussed on delivery for the real world Main architecture, technical implementation and primary capabilities driven by a set of prioritised research questions Based on the main research questions define prioritised data sources Develop three Exemplars to demonstrate the capabilites of the Open PHACTS System and to define interfaces and input/output standards

9. Work Streams Build: Service layer and resource integration Drive: Development of exemplar work packages & Applications Sustain: Community engagement and long-term sustainability

10. Platform Explorer Standards Apps API

11. NumbersumNr of 1Question 15 129 All oxido,reductase inhibitors active <100nM in both human and mouse 18 148 Given compound X, what is its predicted secondary pharmacology? What are the on and off,target safety concerns for a compound? What is the evidence and how reliable is that evidence (journal impact factor, KOL) for findings associated with a compound? 24 138 Given a target find me all actives against that target. Find/predict polypharmacology of actives. Determine ADMET profile of actives. 32 138 For a given interaction profile, give me compounds similar to it. 37 138 The current Factor Xa lead series is characterised by substructure X. Retrieve all bioactivity data in serine protease assays for molecules that contain substructure X. 38 138 Retrieve all experimental and clinical data for a given list of compounds defined by their chemical structure (with options to match stereochemistry or not). 41 138 A project is considering Protein Kinase C Alpha (PRKCA) as a target. What are all the compounds known to modulate the target directly? What are the compounds that may modulate the target directly? i.e. return all cmpds active in assays where the resolution is at least at the level of the target family (i.e. PKC) both from structured assay databases and the literature. 44 138 Give me all active compounds on a given target with the relevant assay data 46 138 Give me the compound(s) which hit most specifically the multiple targets in a given pathway (disease) 59 148 Identify all known protein-protein interaction inhibitors Prioritised research questions Kamal Azzaoui et al, DDT in press 2013

12. ` ` Pathways Pharmacological Activities Biological Processes Transcripts Pathological Processes Diseases Genes Proteins Interactions Clinical Drug Applications Indications Drugs Compounds Chemicals

13. Open PHACTS will be built upon semantic technologies and standards, providing an opportunity to: Demonstrate that semantic technologies can perform to the same degree as existing systems Provide an open platform to address common drug discovery questions; expose pharma’s use-cases and knowledge Create a pre-competitive infrastructure that can be sustained and expanded into new areas; providing the platform for future collaboration Why Semantic Technologies? Rapidly developing technology, powerful algorithms for integration and querying of data “schema free” Open standards – facilitating sharing public, private, commercial A community of developers, leverage work going on elsewhere

14. User Interfaces & Applications Linked Data API Linked Data Cache Identity Mapping Service Identity Mapping Service Identity Resolution Service Domain Specific Services Domain Specific Services Data Key architecture components

15. Nanopub Db VoID Data Cache (Triple Store) Data Cache (Triple Store) Semantic Workflow Engine (LARKC) Linked Data API (RDF/XML, TTL, JSON) Domain Specific Services Open PHACTS Explorer 1 st Gen Apps App Framework Identity Resolution Service (ConceptWiki) Chemistry Normalisation & Q/C ChemSpider Identifier Management Service (BridgeDb+) Partner Apps Data Import Core Platform P12374 EC2.43.4 CS4532 “Adenosine receptor 2a” Oct. 2012 VoID Db Nanopub Db VoID Db VoID Nanopub VoID Public Content Commercial Public Ontologies User Annotations

16. Building Quality High quality chemical names and synonyms. Leverage ChemSpider and Concept wiki curation, Q/C and mapping ChemSpider Validation and Standardization Platform (CVSP) for flagging chemical representation issues Basic curation interface for editing concept terms available through Concept Wiki Data quality issues detected in data sources reported back to depositors for their evaluation

17. STANDARD_TYPE UNIT_COUNT ---------------- ------- AC50 7 Activity 421 EC50 39 IC50 46 ID50 42 Ki 23 Log IC50 4 Log Ki 7 Potency 11 log IC50 0 STANDARD_TYPE STANDARD_UNITS COUNT(*) ------------------ ------------------ -------- IC50 nM 829448 IC50 ug.mL-1 41000 IC50 38521 IC50 ug/ml 2038 IC50 ug ml-1 509 IC50 mg kg-1 295 IC50 molar ratio 178 IC50 ug 117 IC50 % 113 IC50 uM well-1 52 IC50 p.p.m. 51 IC50 ppm 36 IC50 uM-1 25 IC50 nM kg-1 25 IC50 milliequivalent 22 IC50 kJ m-2 20 ~ 100 units >5000 types Implemented using the Quantities, Dimension, Units, Types Ontology (http://www.qudt.org/) Quantitative Data Challenges

18. Chemistry within Open PHACTS The challenges associated with handling chemistry data require the support of a publicly accessible platform to integrate, standardise and host the data. ChemSpider, an online database from the Royal Society of Chemistry hosts the chemical compound collection underpinning Open PHACTS and is responsible for standardising the chemical compounds and providing both regular updates and ongoing data curation. To serve the Open PHACTS platform, a structure validation and standardisation platform (CVSP) has been developed to ensure chemical structures are normalised to rules derived from the FDA structure standardisation guidelines and modified based on input from the EFPIA members.

19. The many challenges of chemistry representation…

20. Identities within Open PHACTS Open PHACTS integrates information from multiple different databases, many of which use unique identifiers. The Identity Mapping Service (IMS) ensures these identifiers are linked and available for use interchangeably throughout the Open PHACTS platform. To maintain vocabulary heterogeneity and provide interoperability, the ConceptWiki is used. The ConceptWiki is an open access system that accepts essentially unlimited numbers of synonyms, in multiple languages, and then maps all the terms correctly back to one unique concept identifier, alleviating vocabulary problems and identifier differences. Synonyms: Aspirin Dispril 2-Acetoxybenzoic acid Acetyl salicylic acid Salicylic acid, acetyl- ChemSpider ID: 2157 Explorer FDA: 16030 ChEBI ID: CHEBI:15365 DrugBank ID: APRD00264 IMS

21. Why Provenance Matters Using a community specification known as “VoID” (Vocabulary of Interlinked Datasets) Record version, author, derivations Builds trust with users – know what you are querying (and why it might have changed) Provides mechanism to provide usage statistics back to providers, help them understand the value Easier to track errors and ensure quality Actively participating in community provenance programme (W3C)

22. What does Open PHACTS do?Currently integrated databases Database Number of triples (million) ACD Labs / ChemSpider 161.34 ChEBI0.91 ChEMBL_v13146.08 ConceptWiki3.74 DrugBank0.52 Enzyme0.07 Gene Ontology0.85 SwissProt156.57 WikiPathways0.14 TOTAL 470.21 Open PHACTS draws together multiple sources of publicly- available pharmacological and chemical data, allowing public access to the information via the Open PHACTS Explorer, an intuitive interface.

23. Licensing: 3 “public” databases Comparative Toxicogenomics Database OMIM Drugbank All are available as “open” RDF you can download right now. But:

24. “CUTTING THE GORDIAN KNOT” What are the problems with licensing we had to address? –To make the data and software generated by the project usable and reusable –Multiplicity of unclear or non-standard licenses on original data sources ‘Public’ can mean use but not redistribute, use in commercial environment, Legal position on use and reuse extremely unclear Different issues than just linking to data –What is the legal status of integrated collections of the above, and of derived knowledge from such a collection? –Appropriate software license selection –Legal clarity for EFPIA and end users –Approaches for commercial data integration, EFPIA in-house data AIM: to enable maximum possible dissemination and usability of the integrated data and architecture generated by the project - with approaches that will be applicable in other data integration projects

25. Chose John Wilbanks as consultant A framework built around STANDARD well-understood Creative Commons licences – and how they interoperate Deal with the problems by: Interoperable licences Appropriate terms Declare expectations to users and data publishers One size won‘t fit all requirements Data Licensing Solution

26. Development partnerships Influence on API developments Opportunities to demo ideas & use cases to core team Need MoU and annexe Associated partners Support, information Exchange of ideas, data, technology Opportunities to demo at ctions, mostommunity webinars Need MoU Associated partners Development partnerships Consortium MoU +Annexe Consortium 28 current members Open PHACTS and the scientific community

27. Example applications Advanced analytics ChemBioNavigator Navigating at the interface of chemical and biological data with sorting and plotting options TargetDossier Interconnecting Open PHACTS with multiple target centric services. Exploring target similarity using diverse criteria PharmaTrek Interactive Polypharmacology space of experimental annotations UTOPIA Semantic enrichment of scientific PDFs Predictions GARFIELD Prediction of target pharmacology based on the Similar Ensemble Approach eTOX collector Automatic extraction of data for building predictive toxicology models in eTOX project

28. ChemBioNavigator Matthias Rarey et al PharmaTrek Jordi Mestres et al

29. Call for expressions of interest Open PHACTS ENSO proposal Open PHACTS intends to submit a proposal for IMI ENSO funding. We are currently drafting our ENSO proposal and invite all EFPIA companies with an interest in Open PHACTS to contact us to discuss opportunities for involvement. The Open PHACTS Foundation Open PHACTS has a successor organisation, the Open PHACTS Foundation. Please register your interest with us for further information on membership and other opportunities to get involved within Open PHACTS. For more information and/or to register interest email us at pmu@openphacts.org pmu@openphacts.org

30. Acknowledgements Stefan Senger Gerhard Ecker The OpenPHACTS consortium

32. Data Targets; Chemistry; Pharmacology; Literature; Patents Standards Ontology/taxonomy; Minimum information guide; Dictionaries; Interchange mapping Assertions e.g. Gene-to-Disease; Compound-to-Target; Compound-to-ADR Application(Knowledge) Fact Visualisation e.g. Target Dossiers; SAR VisualisationSERVICES After Barnes et al Nature Review Drug Discovery 2009 doi10.1038/nrd2944

33. Nanopublications – Capturing scientific information in the Triple Store