Tox 21 Tox Cast Chemical Landscapes Laying the

Tox 21 & Tox. Cast Chemical Landscapes: Laying the Foundation for 21 st Century Toxicology August 25, 2011 5 th Meeting on U. S. Government Chemical Databases & Open Chemistry, Frederick, MD Office of Research and Development National Center for Computational Toxicology Ann Richard National Center for Computational Toxicology

NCCT’s Tox. Cast Program Structures HTS In Vitro QSAR In Vivo Office of Research and Development National Center for Computational Toxicology 1

NCCT Public Data http: //www. epa. gov/ncct/ ACTo. R DSSTox Generic Chemicals Tox. Ref. DB Tox. Cast Test Substances Tox 21 Office of Research and Development National Center for Computational Toxicology

NCCT Public Data http: //www. epa. gov/ncct/ ACTo. R Tox. Ref. DB DSSTox Chemical Structure Annotation Chemical Management Chemical QC Tox. Cast Tox 21 Office of Research and Development National Center for Computational Toxicology

NCCT Public Data http: //www. epa. gov/ncct/ ACTo. R DSSTox Data Integration Chemical – Assay Linkage In vivo data Linkage Tox. Ref. DB Analysis tools Tox. Cast Tox 21 Office of Research and Development National Center for Computational Toxicology

EPA NCCT launches Tox. Cast Office of Research and Development National Center for Computational Toxicology is be g tin g 2009 Phase II: (700) To x Ph as e It es Phase I (309) 2008 NAS Report released Ca st D Su ata m A m na it lys ins tra cts t. P as x. C 2007 To To x Ca st co nc h iss ase ue I c d on ep tp ap er To x Sc i Tox. Cast Project Phase I data results ACTo. R launched 2010 Full Phase I data publication/release Tox. Ref. DB web access w/in ACTo. R

Tox 21 Project 10 K 5. 6 K 2. 8 K 1. 4 K 2004 2006 Tox 21 Chemical Library 2010 2008 2007 2005 NTP HTS Plate A Tox. Cast Phase I NCGC Drug Plates A, B EPA Tox 21 Plate A 2009 2011 Tox. Cast Phase II Tox 21 EPA Plates A’, B, C NTP Tox 21 Plates A’, B, C NCGC Drug Plates A’, B’, C Office of Research and Development National Center for Computational Toxicology

Tox 21 Chemical x Assay Landscape Tox 21 NTP HTS Plate A Tox. Cast Phase I & II ~10, 000 Chem x 50 -100 assays 1408 Chem x >100 assays 960 Chem x >500 assays 100 # Assays >500 ~10, 000 # Chemicals 1536 well microplate format (1408 cmpds/plate) x 9 plates NCCT/EPA Environmental Industrial NIEHS/NTP Pesticides Food Use Drugs Toxicology FDA Drugs NIH/NCGC CFSAN/CDER Office of Research and Development National Center for Computational Toxicology 7

Tox 21: EPA Chemical Inventories TOX 21_EPA_3726 e 1 k Tox. Cast_ 960 100 820 unique (1000 samples) To be tested in endocrine-related subset of Tox. Cast assays IIc Phase I & II(a, b) 111 (+24 Ph. IIc) failed drugs, CFSAN, pesticides, EPA high interest compounds tested in >500 HTS assays Office of Research and Development National Center for Computational Toxicology

Tox 21: EPA Tox. Cast Chemicals Tox. Cast Phase I (293 unique) EPA pesticidal actives w/ rich in vivo data Ø PFOAs, BPA, metabolite/parent pairs Ø Tox. Cast Phase II a, b (776) +c (100) EPA pesticides, high interest EPA and stakeholder inventories, data rich chemicals (EDSP, OPPT, Antimicrobials, Inerts, …) Ø 135 Pharma failed drugs w/ pre-clinical & clinical tox data Ø FDA CFSAN data rich, NCTR LTKB Priority 1 drugs Ø Toxicity reference chemicals, data-rich chemicals, NTP immunotox Ø Broaden selections from Tox. Cast Phase II nomination lists to include in Tox 21 EPA set Office of Research and Development National Center for Computational Toxicology

e. g. , EPA Tox 21 Chemical Procurement HPV, MPV Substances NCCT Programs EU, FDA Endocrine disruptors Water contaminants In vivo data availability (EPA, NTP, FDA) Antimicrobials Pesticide actives Pesticide inerts Industrial chemicals Green Chemistry EPA Tox 21/Tox. Cast Phase II Chemical Nominations (over 100 lists) ~19, 000 compounds Candidates for procurement ~7, 000 compounds Unable to procure Too expensive Able to purchase 4371 compounds COA/MSDS review Solubilize in DMSO 211 volatile/ insufficient sample Complex mixtures Ill-defined substances No structure Insoluble (est. Log. P) Volatiles (est. VP) Inorganics Explosive Reactive Polymers … 435 Insoluble in DMSO (approx 10%) EPA Office Tox 21 3726 unique of Research and Development National Center for Computational Toxicology Inventory substances 10

Tox 21 Chemicals: EPA Selection Strategy Nominations & tracking of inventory overlaps by CAS Prioritize for procurement Cost & availability Filter by MW, physchem, volatility, MSDS cautions DMSO solubility Office of Research and Development National Center for Computational Toxicology

Tox. Cast Ph. I&Ph. II 960: # Compounds per Inventory (960 x 16) Series 1, AIR, Series 1, GRAS, 90 26 Series 1, EDSP, 120 Se rie Series 1 s 1, , TRI, Series 1, MPV, IRI 216 83 Seri S, es 1, Series 1, Green Chemistry, 24 HP 75 Series 1, Antimicrobials, 0 V, 91 Series 1, 232 Consumer, Series 1881, S Water, er 217 ie s 1 , P es tic id Office of Research and Development e. I National Center for Computational Toxicology n er 960 Total chemicals 2740 total overlaps across 16 diverse inventories Excellent coverage of multiple high-interest inventories Broad diversity of chemicaluse and types Large overlap with data-rich inventories 12

Tox. Cast_960 Pharmaceuticals: Multipurposing (255) 111 Failed pharmaceuticals – no overlaps 144 additional cmpds classified as drugs; on other lists with other uses 137 of these appear on 2 or more lists Caffeine appears on 18 lists Office of Research and Development National Center for Computational Toxicology 13

Chemical Sample Registration Workflow Place chemical orders Partition to Matrix racks Procure 2 x 100 mg compound in tared, barcoded vials Store + Solubilize neat in DMSO Solution Plates Assay Providers Plate ID NCGC Plate Address Analytical QC Review COA Bottle IDs X Insoluble QC structure Register in DSSTox Inventory Solution IDs Analytical QC Register in Sample Tracking Database Update Assay Results ACTo. R/ Pub. Chem Update Remove from DSSTox Office of Research and Development National Center for Computational Toxicology 14

Tox 21 Analytical QC A copy of each parent Tox 21 assay plate (352 cmpds/plate) will be subjected to analytical QC for assessing purity, identity, stability LC-MS PASS = Confirm parent ion peak and >90% purity Fail, inconclusive or analytical method inappropriate Retest at later time point under assay conditions for stability Publish QC summary results in association with assay data Office of Research and Development National Center for Computational Toxicology GC-MS 15

Analytical QC Summary PDF results will be available for each sample NCGC Analytical Chemist (W. Leister) will review all preliminary LC results and QC “FAIL” compounds, and supervise follow-up testing Prime objective of QC is to inform analysis & interpretation of assay results: high confidence low confidence fail Office of Research and Development National Center for Computational Toxicology

Tox. Cast/Tox 21 Chemical Registry Analog Searching External Resources Structure-based Data Mining AIM DSSTox Pub. Chem. Spider EPA ACTo. R e DSSTox tur Chemical structure - CID Substance details - SID Project inventory record - RID ru St c SAR Modeling Assay Results CID Table Structures Tox 21 Tox. Cast Tox. Ref. DB Test Sample RID Table Office of Research and Development National Center for Computational Toxicology 7

Tox. Cast/Tox 21 Chemical Registry Bottle_ID DSSTox_RID Chemical structure - CID Substance details - SID Project inventory record - RID COA_ID Assay Results QC_ID Soln_ID Structures Test Sample Office of Research and Development National Center for Computational Toxicology DSSTox RID Bottle ID ( COA ID) Solution ID ( QC ID) Tox 21 Sample Tracking Database 8

Tox. Cast/Tox 21 Chemical Registry Pub. Chem_SID Pub. Chem_CID ACTo. R/Tox. Miner tur e DSSTox St Chemical structure - CID Substance details - SID Project inventory record - RID r DSSTox RID uc Structures Tox 21_ID Assay Results Assay name Assay details Assay outcome Solution ID Plate Address ID Test Sample Office of Research and Development National Center for Computational Toxicology DSSTox RID Bottle ID ( COA ID) Solution ID ( QC ID) Tox 21 Sample Tracking Database 9

Chemical Information & Sample QC “Whoever is careless with the truth in small matters cannot be trusted with important matters. ” – Albert Einstein

Gene-expression Chemical Sample Details Chemical Annotation Layer Chemical Supplier Lot/Batch Certificate of Analysis Solubility Purity & Stability Chemical Name CAS SMILES In. Ch. I / Structure Substance details Toxicology literature Public databases Cheminformatics resources

Chemical Information & Sample QC Chemical Annotation of Public resources Generic Chemical of Toxicologica l Interest ü Name is misspelled or incorrect ü CAS is invalid or retired ü CAS and name do not agree ü Name and structure do not agree ü Name is insufficient for structure assignment ü Insufficient description of substance Sample Annotation & QC Procure from Chemical Supplier Name, CAS, purity (COA) MW, dose IC 50 ü No CAS or wrong ü Name wrong ü Structure wrong ü Not same as generic chemical Salt, isomer, … ü Name does not match info on COA hydrate, stereo ü Wrong MW ü COA expired ü Purity <90% ü Active impurities ü Sample degrades ü Rxn with solvent ü Hydrolysis

Tox 21 Cheminformatics CAS/Name Creating SAR-ready files: How to process salts, complexes, charged species How to aggregate results for resulting structure “duplicates” How to deal with enriched features & “families” of structures within dataset Chemical Substance: SID 2 D, 3 D Conformers Tautomers Mixture IDs Structure: CID Salts Complexes (Metabolites? ) Parent: PID Tox 21/Tox. Cast Assay Results Groups: Map. ID Features Classes Descriptors Properties Analogs, classifiers, stereo family, etc. Bottle IDs Analytical QC Solution IDs Plate IDs, Plate Addresses Office of Research and Development National Center for Computational Toxicology 23

DSSTox TOX 21 E 1_3619 13 macromolecule 155 mixture or formulation 1 unspecified or multiple forms 45 inorganics 100 organometallics 173 complexes 219 salts 12 no structures Alkylbenzenes Chlorobenzenes Methyl phenols Nitrobenzenes Aromatic amines Phthalates Perfluorinates… NCGC Pharmaceuticals NTP compounds FDA compounds EPA compounds 91 “parent” groupings (2 -3) MW ranges from 30 (formalin) to 1700 (tannic acid) Office of Research and Development National Center for Computational Toxicology 24

Tox 21 Structural Library Feature enrichment Feature diversity Office of Research and Development National Center for Computational Toxicology

Tox 21 Reaction Features: Commonalities NTP Tox 21 NCGC Tox 21 EPA Tox 21 Drugs Food additives Antimicrobials Water contaminants HPVs Toxicants … Office of Research and Development National Center for Computational Toxicology Metabogen reaction features generated using “MOSES” software by Molecular Networks

Tox 21: Molecular Weight Distributions Tox 21 NTP & EPA # Molecules Mean = 231 Mean = 304 Tox 21 NCGC Molecular Weight Office of Research and Development National Center for Computational Toxicology NTP & EPA collections enriched with lower MW compounds compared to NCGC drug collection

Tox. Cast/Tox 21 property distributions Tox. Cast_Phase. I LOG P = Octanol/Water partition coefficient TPSA = log (Total Polar Surface Area) Complexity = log (complexity based on paths, branching, atoms) Office of Research and Development National Center for Computational Toxicology Chemical properties computed using “Adrianna” software by Molecular Networks.

Tox. Cast/Tox 21 property distributions Tox. Cast_Phase. II LOG P = Octanol/Water partition coefficient TPSA = log (Total Polar Surface Area) Complexity = log (complexity based on paths, branching, atoms) Office of Research and Development National Center for Computational Toxicology Chemical properties computed using “Adrianna” software by Molecular Networks.

Tox. Cast/Tox 21 property distributions Tox. Cast_Phase. II 111 failed drugs LOG P = Octanol/Water partition coefficient TPSA = log (Total Polar Surface Area) Complexity = log (complexity based on paths, branching, atoms) Office of Research and Development National Center for Computational Toxicology Chemical properties computed using “Adrianna” software by Molecular Networks.

Tox. Cast/Tox 21 property distributions Tox. Cast_Phase. II 111 failed drugs Tox 21 LOG P = Octanol/Water partition coefficient TPSA = log (Total Polar Surface Area) Complexity = log (complexity based on paths, branching, atoms) Office of Research and Development National Center for Computational Toxicology Chemical properties computed using “Adrianna” software by Molecular Networks.

Expanding Applicability QSAR Defined organics Virtual chemicals Volatiles DMSO insolubles Reactives In vitro/HTS Defined structure + Biological profile metabolism ADME In Vivo Office of Research and Development National Center for Computational Toxicology Mixtures Formulations Proprietary substances Organometallics Metals Salts

Acknowledgements: EPA NCCT Tox. Cast Team: Robert Kavlock - Director David Dix Keith Houck Matt Martin (Tox. Ref. DB) Richard Judson (ACTo. R) EPA NCCT DSSTox: Maritja Wolf– Lockheed Martin, Contractor to the EPA Indira Thillainadarajah – EPA: SEE Patra Volarath – EPA Post Doc External Collaborators: Chihae Yang, FDA/CFSAN Chris Austin & colleagues, NCGC/NIH Ray Tice & colleagues, NTP/NIEHS Office of Research and Development National Center for Computational Toxicology This work was reviewed by EPA and approved for publication but does not necessarily reflect official Agency policy.