Note: this document may contain some
elements that are not fully accessible to
users with disabilities. If you need
assistance accessing any
information in this document, please
contact ORD_Webmaster@epa.gov.
US EPA CSS-HERA
Board of
Scientific
Counselors
Chemical Safety
Subcommittee
Meeting
US EPA CSS-HERA BOSC Meeting - February 2-5, 2021
SW(
$
<
3J
\
SB
ro
z
111
p*
PR0^°
The work presented within represents US EPA Office of Research and Development research
activities. Material includes both peer reviewed, published results and work-in-progress
research. Please do not cite or quote slides.
-------
Table of Contents
Approaches and Models for Species Extrapolation (Carlie LaLone) 3
Novel in vitro Methods for Ecological Species: Evaluating Cross-species Differences in Nuclear Receptor-
Ligand Interactions (Brett Blackwell) 30
High Throughput Transcriptomics: A Multi-Species Approach (Kevin Flynn) 50
The work presented within represents US EPA Office of Research and Development research activities. Material
includes both peer reviewed, published results and work-in-progress research. Please do not cite or quote slides.
-------
The views expressed in this poster are those of the authors and do not necessarily reflect the views or policies of the U.S. EPA.
&EPA
United States
Environmental Protection
Agency
Approaches and Models for Species
Extrapolation
Carlie A. LaLone, Ph.D.
Research Bioinformaticist
Office of Research and Development
Center for Computational Toxicology and Exposure, Great Lakes Toxicology and Ecology Division
February 2nd, 2021
Cross Species Extrapolation
TOXICOKINETICS
Absorption
TOXICODYNAMICS
Elimination
Distribution
Metabolism
Bioinformatics
-------
A EPA
United States
Environmental Protection
Agency
Partners in Regulatory Decision-Making
• Capitalize on available data
Clean Air Act
Clean Water Act
Resource Recovery Act
Endangered Species Act
Food Quality Protection Act
Endocrine Disruptor Screening Program
Federal Insecticide, Fungicide, and Rodenticide Act
Frank R. Lautenberg Chemical Safety for the 21st Century Act
Comprehensive Environmental Response, Compensation, and Liability Act
Guidelines for Deriving Numerical National Water Quality Criteria for the Protection of Aquatic Organisms and Their Uses
Confidence in decisions, with limited data available, limited resources for testing, strong backing to reduce animal
-------
A EPA
United States
Environmental Protection
Agency
Stakeholder Identified Challenges
• Limited or no toxicological data for the animal or plant species of
interest - reliance on surrogate (model organisms)
• Impractical to generate new data for all species
• Testing resources are limited
• EPA directive aligns with international interest to reduce animal use
• Ever-increasing demand to evaluate more chemicals in a timely and sometimes
expedited manner
• Sensitivity of species must be estimated based on scientifically-sound
methods of cross-species extrapolation
• Immense diversity of species in the wild
• Important challenge for species listed under the Endangered Species Act
3
-------
-------
AEPA
United States
Environmental Protection
Agency
Chemical Safety Evaluation
• Protect human health and the environment
• Ensure that chemicals in the marketplace are reviewed for safety
• Challenging mission:
• Tens of thousand of chemicals are currently in use and hundreds are
introduced annually
• Many have not been thoroughly evaluated for potential risk to human health
and the environment
• Chemicals tested across species: Even more sparse
-------
6
-------
a rpA
Strategic Approach to Species Extrapolation
United States
Environmental Protection
Agency
Computational:
Bioinformatics (Session 2 Demo)
Systematic review
Case Examples:
PEAS targets
Endocrine pathways
Pollinators
Experimental:
Site-directed mutagenesis
Attagene XS-2 Factorial assay (Dr. Blackwell)
-------
Predictive
\X 4T"
Approache:
i Sequence Structure
Function
-------
https://seqapass.epa.gov/seqapass/
Sequence Alignment to
Predict Across Species
Susceptibility
(SeqAPASS) •
s
Cf\ I 1 Society of
OV_/ 1 Toxicology
www.toxsci.oxfordjournals.org
TOXICOLOGICAI SCIENCES, 153(2), 2016, 22S-245
dot: 10.1093/toxsci/kfwl 19
Advance Access Publication Date: June 30,2016
Research article
Sequence Alignment to Predict Across Species
Susceptibility (SeqAPASS): A Web-Based Tool for
Addressing the Challenges of Cross-Species
Extrapolation of Chemical Toxicity
Carlie A. LaLone *,:L Daniel L. Villeneuve,* David Lyons,f Henry W. Helgen,*
Serina L. Robinson,5,2 Joseph A. Swintek,11 Travis W. Saari,* and
Gerald T. Ankley*
CVJSTO/i^
BUILT
^Or YO^
-------
What information is required for a SeqAPASS query?
United States
Environmental Protection
Agency
1. Protein
2. Species
Chemical-Protein Interaction
Knowledge of a sensitive or targeted species
Knowledge of the model organism used in an m vitro assay
Knowledge of the species for which the Key Event was developed
Chemical Molecular Target
in Target Species
Compare to Millions of Proteins
From housands of Species
Greater similarity = Greater likelihood that chemical can act on the protein
Line of Evidence: Predict Potential Chemical Susceptibility Across Species
10
-------
Flexible Analysis Based On Available Data
Level 1
Primary Amino Acid Sequence Alignments
Level 2
Conserved Functional Domain Alignments
Level 3
Critical (Close Contact) Amino Acid Conservation
seqapass.epa.gov/seqapass/
/ s
A
•* t
4
» V
4
^ J
./
./yv
./ yv*
./ Cn,-v
[ >
n
./ yv-
Gather Lines of Evidence Toward Protein Conservation
-------
SeqAPASS Predicts Likelihood of Similar
Susceptibility based on Sequence Conservation:
W
* no
no
Line(s) of evidence indicate
• The protein is conserved
• The protein is NOT conserved
-------
oE
United States
Environmental Protection
Agency
Evolution of the SeqAPASS tool
v5.0 (Nov. 2020): Develop visualization (Level 3), Develop Decision
Summary Report
v4.0 (2019): Improve visualization, user guidance, summary tables,
interoperability
v3.0 (2018): Develop visualization (Level 1 & 2), automate Level 3
Susceptibility Predictions
v2.0 (2017): develop Level 3 Susceptibility Predictions
vl.O (2016): Develop interface Level 1 & 2 and integrate essential
functionality
CRAWL
#1 ^ WALK
L. V RUN
-------
a rpA
Strategic Approach to Species Extrapolation
United States
Environmental Protection
Agency
Computational:
Bioinformatics (Session 2 Demo)
Systematic review
Case Examples:
PEAS targets
Endocrine pathways
Pollinators
Experimental:
Site-directed mutagenesis
Attagene XS-2 Factorial assay (Dr. Blackwell)
-------
Deiodinase 3: Important enzyme in thyroid function
t
Partial Match Susceptible Yes
Ml
lodothyronine Deiodination
Tetra-iodothyronine (Thyroid
Hormone T4)
Site Directed Mutagenesis to Probe SeqAPASS Level 3
Mutate hDI03
Not a Match Susceptible No
SeqAPASS Critical Amino Acid Comparison
hDI03 mutants:
Sea lamprey
Fish
Fish
Frog
Lungfish
Sea lamprey
C168G
T169S
C239S
A240R
Y257A
Y257F
Common Name
Similar Susceptibility
| Amino Acid 1 |
| Amino Acid 2 |
| Amino Acid 3 | Amino Acid 4 |
| Amino Acid 5 |
Human
¦¦
Red drum
201S
202N
Blue tilapia
195L
Tongue sole
202N
Zebra mbuna
1 i3is 1
201T
202N
Nile tilapia
201T
202N
Senegalese sole
1 1315 1
201G
202N
Ballan wrasse
1 1315 1
201S
202N
Gilthead seabream
1 1315 1
201S
202N
Monterrey platyfish
196R
Amazon molly
196R
Shortfin molly
196R
Sapphire devil
130S
200G
201N
Southern platyfish
196R
Goldlined spinefoot
202E
Torafugu
202N
Sailfin molly
202N
Threespot wrasse
182N
Princess parrotfish
169N
Striped parrotfish
157N
1 - 1
Frogs and toads
204R
Two-lined caecilian
1 - 1
t29C
1301 ]
20 IP
Puerto Rican coqui
201C |
202R
Caecilians
09C
BOT
200C ~]
201P
Tropical clawed frog
13 IT
' 201C '
202R
African clawed frog
L29T
I99C ]
200R
Gabon caecilian
1 r |
131T
2G1C J
202P
American bullfrog
-0-C
203R
Coelacanth
d 1
202C
203F
Sea lamprey
1 1430 1
| I44S |
2I6P
1 233A I
Human Deiodinase 3 Enzyme
-------
Strategic Approach to Species Extrapolation
Computational:
Bioinformatics (Session 2 Demo)
Systematic review
Case Examples:
PEAS targets
Endocrine pathways
Pollinators
Experimental:
Site-directed mutagenesis
Attagene XS-2 Factorial assay (Dr. Blackwell)
-------
vvEPA
U.S. EPA Toxicity Forecaster ('l bxCast)
U.S. EPA ToxCast Program:
US EPA ToxCast Program: Uses mammalian cell-based assays to rapidly screen chemicals, identify putative molecular
targets, and identify those most likely to be endocrine disruptors
%
Key Questions for Consideration:
• How well does this mammalian-based prioritization approach reasonably reflect potential impacts on other
vertebrates?
• Can we expect chemicals that interact with mammalian receptors to also interact with receptors of other speciesf 7
-------
vvEPA
Hierarchal Framework for Evaluating Pathway
Conservation Using Existing Evidence
Analysis of structure
across species using in
silico computational tools
$trytfural Conservation (Molewlar Initiating Event)
¦ In silico approaches (e.g.. for proteins)
• Primary ammo a Environ Toxicol Chern. 2016 Nov;35(11):2806-2816. doi: 10.1002/etc.3456. Epub 2016 Jun 28,
Evaluation of the scientific underpinnings for
identifying estrogenic chemicals in nonmammalian
taxa using mammalian test systems
Gerald T Ankley 1, Carlie A LaLone 2, L Earl Gray 3, Daniel L Villeneuve 2, Michael W Hornung 1
What other important endocrine targets
have a large base of pre-existing structural,
molecular target, and toxicity data?
-> Androgen Receptor (AR)
18
-------
Assess?ngAR Conservation Across Species Using the SeqAPASSTool
Invertebrates
Level 1: Analysis of AR Primary
Amino Acid Sequence
910 species
evaluated
A
1 if II III I 1| I I I I lit
fUi lit J!lmnIiia8fsiUrrn! Hfi
" " !i " " J S J
s i I =? =
S S
Vertebrates
I i u i
m a
Species Taxonomic Class
I 1
Invertebrates
Level 2: AR Ligand Binding
Domain
907 species
•
- • evaluated
i
1
1
1
•
I
• m
1
t
• II
»
1
0 1
¦
|j. Aw
a
1
~ • ~ A
infill f
• I Ml
s *
fWipPPWIW
HIT
H
3.
Level 3: Analysis of Conservation of
Individual Amino Acid Residues
250 species evaluated
Taxonomic Group
# of Spp.
Shared
Susceptibility
Mammals
117/1
Yes/No
Lizards, Snakes
11
Yes
Turtles
3
Yes
Birds
58
Yes
Crocodiles,
A
VaC
Alligators
i es
Amphibians
13
Yes
Coelacanths
2
Yes
Eel-shaped
1
Yes
Bony Fish
87/1
Yes/No
Sharks, Rays
4
Yes
Lungfish
2
Yes
Across all three levels, SeqAPASS results suggest conservation of
AR across vertebrate species
Overall, these predictions suggest that chemicals that bind and
activate AR in mammalian-based assays, are likely to interfere with
AR in other vertebrate species
Line of evidence for pathway conservation
Species Taxonomic Class
19
-------
Evaluating Existing Data to Extrapolate High-Throughput Androgen
Receptor Screening Data Across Species
Fewer Resources
*
Structural Conservation (Molecular Initiating Event)
¦ In silico approaches (e.g., for proteins)
• Primary amino acid
• Functional domain
• Individual residues involved in chemical binding
Less Certainty
t
- I
Functional Conservation (Cellular Response)
¦ In vitro assays
• Competitive binding assays
• Transcriptional activation assays
I
Comparative Analyses (Organism Response)
In vivo studies
• Apical responses to chemicals
and adverse outcomes
More Resources
Greater Certainty
Evidence for Pathway Conservation
> Environ Toxicol Chem. 2016 Nov;35(11}:2806-2816. doi: 10.1002/etc.3456. Epub 2016 Jun 28.
Evaluation of the scientific underpinnings for
identifying estrogenic chemicals in nonmammalian
taxa using mammalian test systems
Gerald T Ankley 1, Carlie A LaLone 2, L Earl Gray 3, Daniel L Villeneuve 2, Michael W Hornung 2
¦I
K
Systematic Literature Review: A type of literature review that
uses systematic methods to collect secondary data, critically
appraise research studies, and synthesize findings
Using existing evidence (literature), we can evaluate the
scientific basis of our cross-species predictions
Advances in data science can improve this workflow
Gathering in vivo and in vitro data from vertebrate species
exposed to known androgenic compounds provides
additional lines of evidence for the conservation of the
biological pathway across species
20
-------
Evaluating Existing Data to Extrapolate High-Throughput Androgen Receptor
Screening Data Across Species
SeqAPASS Evaluation of Structural
Conservation Across Species
Systematic Evaluation of In Vitro
Cross-Species Data
Conserved?
Yes
No
Systematic Evaluation of In Vivo
^ Cross-Species Data
Weight of Evidence for Pathway
Conservation Across Species for
Defined Risk Assessment
Applications
Apply pathway to other targets of interest
Repeat process to account for the emergence of new information
21
-------
Strategic Approach to Species Extrapolation
Computational:
Bioinformatics (Session 2 Demo)
Systematic review
Case Examples:
PEAS targets
Endocrine pathways
Pollinators
Experimental:
Site-directed mutagenesis
Attagene XS-2 Factorial assay (Dr. Blackwell)
-------
I
A EPA
United States
Environmental Protection
Agency
Sequence
MTMTLHFKASGMALLHdllQGNEltPLNRPClLKIf'ttftPLGE
VYLQ55KPAYYNYPEGAA7r£FNAAAAANAQVYGQT(5tPTO
PG 5E AAA F LG G FP P L N S VS PSPl M LLH P P P Q.LS W LQ
PMGQQVPWLE N EPSGYTVREAGPPAjVRPhiSDNfl RQGGR
£RLASTNDKGSMAME^KFTftY£^^fcyM6YH¥G'iWSC
r-..... T..^- - ^. ,.r, *y0F?t'tTr" —\ • •-1 r
KCYtVGMMKGClRXOfA^nKHKRQROOOEOftCEVO
SAGDMRAftNLWPSPLMIRRKKNSLALSnADQMVSALLfl.
EPPILySEVDfTRPFSEASMUQLOTMLADRELVHMINWAKV
PGFVDtUHDQVMLl£CAWLEiLMIGLVWRSMEHPGKLLFA
PNUl DfiNOGKCVEGMVEIFDMLL ATSSfl FRMM NlQG EEF
VaK<.l!H,NS6VYTFlSSn!(Sl£El(OHilHRVlOKIT[>TLIHlM
Structure
Function
O
Improvements
in bioinformatics
Yes or No
Susceptible or Not Susceptible
Structural-based
comparisons of similarity
Predicted binding affinity
23
-------
SERA
United States
Environmental Protection
Agency
Advances in Drug Discovery/Development
Structure derived
from X-ray
crystallography
Human
Protein Structure
Bio in form at ics Toolbox:
Molecular modeling
Molecular docking
Virtual screening
Molecular dynamic simulations
Ligand X
2nd
Ligand Y
Ligand Z
r
1st
24
-------
SERA
United States
Environmental Protection
Agency
Application to Species Extrapolation
Human
Protein Structure
Ligand X
2nd
Frog
Protein Structure
Ligand Y
/
Ligand X
2nd
1st
Turtle
Protein Structure
Ligand X
1st
Ligand Y
Bird
Protein Structure
Ligand X
2nd
Bioinformatics Toolbox:
Molecular modeling
Molecular docking
Virtual screening
Molecular dynamic simulations
1st
Fish
Protein Structure
%
2nd
3rd
Fly
Protein Structure
25
-------
SERA
United States
Environmental Protection
Agency
Sequence
MTMTLHTKASGMALLHQIQGNELEPLNRPQLKIPLERPLGE
VYLDSSKPAVYNYPEGAAYEFNAAAAANAQVYGQTGLPYG
PGSEAAAFGSNGLGGFPPLNSVSPSPLMLLHPPPQLSPFLQ
PHGQQVPYYLENEPSGYTVREAGPPAFYRPNSDNRRQGGR
ERLASTNDKGSMAMESAKETRYCAVCNDYASGYHYGVWSC
EGCKAFFKRSIQGHNDYMCPATNQCTIDKNRRKSCQACRLR
KCYEVGMMKGGIRKDRRGGRMLKHKRQRDDGEGRGEVG
SAGDMRAANLWPSPLMIKRSKKNSLALSLTADQMVSALLA
EPPILYSEYDPTRPFSEASMMGLLTNLADRELVHMINWAKV
PGFVDLTLHDQVHLLECAWLEILMIGLVWRSMEHPGKLLFA
PNLLLDRNQGKCVEGMVEIFDMLLATSSRFRMMNLQGEEF
VCLKSIILLNSGVYTFLSSTLKSLEEKDHIHRVLDKITDTLIHLM
Structure
n | f * r
,/VYYYN
Final model
SeqAPASS Results from Level 1
Qiieiy Sequence FASTA + FASTA from 100s
of Aligned Sequences
Across Taxa
>NP_001434.1 Protein X [Homo sapiens]
MSFSGKYQLQSQENFEAFMKAIGLPEELIQKGKDI
KGVSEIVQNGKHFKFTITAGSKVIQNEFTVGEECE
LETMTG EKVKTWQLEG DN KLVTTFKNIKSVTELN
GDIITNTMTLGDIVFKRISKRI
>NP_787011.1 Protein X [Bos taurus]
MNFSGKYQVQTQENYEAFMKAVGMPDDIIQKGK
DIKGVSEIVQNGKHFKFIITAGSKVIQNEFTLGEECE
MEFMTGEKIKAWQQEGDNKLVTTFKGIKSVTEFN
GDTVTSTMTKGDWFKRVSKRI
>KFQ76585.1 Protein X [Phoenicopterus ruber
ruber]
MSFTGKYELQSQENFEPFMKALGLPDDQIQKGKD
IKSISEIVQDGKKFKVTVTTGSKVMQNEFTIGEECD
IEMLTGEKVKAWQMEGNNRLVANLKGLKSVTEL
NGDIITHTMTMGDLTYKRISKRI
>NP_001116883.1 Protein X [Xenopus
tropicalis]
MAFAGKYELVHQENFETFMKAIGLSDELIQKGKDV
KSVTEIQQNGKHFIVTVTTGSKVLRNEFTIGEEAE
LETPTG EKVKSWKLEG DNKLWQLKAITSTTELSG
DTITHVLTLNNLVFKRVSKRV
- 100s of FASTA
Sequences
Predicting Binding Affinity
Ligands of Interest for Docking
Collect Predicted Binding Affinity
re-assembly
Protein Structure Models
From 100s of Species
Generate
Protein Structures
For 100s of species
Template Cluster Centroid
Graphic Modified from Zhang et al„ 2019 1-TASSER gateway: A protein structure and fimction prediction server powered by XSEDE Figure 1
Iterative Threading ASSEmbly Refinement
Develop Models for 100s of Species Based on Aligned Sequences
(I-TASSER; https: zhanglab.ccnib.med.umich.edu/I-TASSER/)
UCSF Chimera
DockPrep Structures and Minimize Ligands
AutoDock Vina
Dock Multiple Ligands to Protein Structures
assembly
Structure
-------
A EPA
United States
Environmental Protection
Agency
Acknowledgements
U.S. EPA. ORD
Donovan Blatz (ORISE)
Sara Vliet (ORISE)
Sally Mayasich (ORISE)
Marissa Jensen (Univ. Minnesota Duluth)
GDIT
Thomas Transue
Cody Simmons
Audrey Wilkinson
SeqAPASS v5.0
https ://seqapass. epa. gov/ seqapass/
27
-------
oEPA
United States
Environmental Protection
Agency
Novel in vitro Methods for Ecological
Species: Evaluating Cross-species Differences
in Nuclear Receptor-Ligand Interactions
B.R. Blackwell, D.L. Villeneuve, G.T. Ankley, CA. La Lone, J.A. Doering
\
Office of Research and Development
Center for Computational Toxicology and Exposure, Great Lakes Toxicology and Ecology Division
Progress for o Stronger Future
-------
W|P_ chemicals in the Environment
• Hundreds of chemicals detected in
contaminant Pathways the environment
• Contaminants of emerging concern
(CECs) include those with limited
toxicity data
• Recognized as critical data gap by
stakeholders:
• OCSPP; OW; Regions: Great Lakes
National Program Office (GLNPO)
• Effects-based monitoring a tool for
understanding impacts and potential
risks of contaminants
t\\ Runoff and
\ Discharges
Sediment
Rivers and
Streams
-------
SERA
United States
Environmental Protection
Agency
nvironmental Monitoring with Attagene TRANS-FACTORIAL Assay
THRal ERa
PXR ~ ' LXRa
ERRa
NURR1
HNF4a
RXRb
PPARa
RARb
RARa
PPARg
LXRb
RORb
RXRa
FXR
PPARdl 8
PXR
ERa
PPARy
GR
PPARa
RXR|3
Do the human receptors
adequately represent
sensitivity of aquatic
vertebrate receptors?
Among the most frequently detected
nuclear receptor activities in surface
water samples
-------
¦SEPA r | .
Cross-species extrapolation
United States
Environmental Protection
Agency
To date, high throughput screening has been human centric
Unclear how well mammalian HTS assays represent
vertebrate diversity, let alone other phyla.
Not feasible to include all taxa in a HTS screening program.
How can we strategically select the minimum number of
representative species that cover the maximal range of
variation in sensitivity and specificity?
4
-------
AEPA
United States
Environmental Protection
Agency
Attagene EcoTox FACTORIAL Assay
t
GAL4-NR1
GAL4 RTU,
9
GAL4-NR2
GAL4RTU, ,
-
•
•
•
GAL4-NR»,
, GAL4 RTUn ,
-+¦
Evaluated
Compound
O
profiling
reporter RNAs
O
1
o
o
1
AR
Reptilian
Chrysemys picta
XM
005279527
AR
Avian
Gaiius gaiius
NM
001040090
AR
Mammalian
Homo Sapiens
NM
000044
TRa
Fish
Danio rerio
NM.
.131396.1
TRP
Danio rerio
NM
131340.1
TRa
Amphibian
Xenopus iaevis
NM
ID
-------
jg™S, Pro,action SpGC iGS S G l G Ct \ O Pi
Is the selection of one representative vertebrate from each class the
best way to cover the potential variability in sensitivity?
Could available information be used to guide a more strategic selection?
• Documented species differences in sensitivity to ligands
• Amino acid residues identified as critical to ligand binding in one
or more species
• In silico analyses of conservation/variation in aa sequence using
SeqAPASS
-------
&EPA
United States
Environmental Protection
Agency
PPARy - established cross-species differences
-
->T _ «v
" W
rmKKKF
|JL
v -¦• V.'V. o
Zebrafish ]
^ Plaice I
Sea Bream ]
Clawed Frog HI
\
A*.
Rat
Mouse
> 200-fold difference In
transactivation of PPARy
by Rosiglitazone
Human
o
0.2
0.4 0.6 0.8 i 1.2 1.4
Relative Transactivation
-------
SERA
United States
Environmental Protection
Agency
SeqAPASS Analysis
https://seqapass.epa.gov/seqapass/
03
LT)
50
c
O)
u 40
Zebrafish
I
Compares amino acid sequence information for
all species for which there are data in the NCBI
protein database.
• Level 1 - Primary Sequence
• Level 2 - Conserved domains
• Level 3 - Individual amino acid residues
T ^
T T
i/i i/> rb */) >•
< -a
o
u
x
_ft
Q.
C O
O -C
ci y
o
2 T3 ^
Taxon
-------
mple SeqAPASS Level 3 - PPARy
• Only 4 positions showed important differences in amino acids among PPARy
• 2 positions known to significantly alter interaction of ligand (rosiglitazone) with PPARy
Taxa
# of Species
Position 1
(Ile309)
Position 2 Position 3
(Gly312) (Cys313)
Position 4
(Tyr355)
Susceptibility
Prediction
Human
1
1
G C
Y
Yes
All Mammals 107
1
G C
Y
Yes
Mallard-type 1
1
G C
Y
Yes
Most Birds 70
1
R C
Y
No
All Reptiles 19
1
R C
Y
No
Strongly conserved among most birds,
All Amphibians 2
1
R C
F
No
amphibians, reptiles
Ancient Fishes 9
F
R C
1
No
Most Fishes 61
F
S C
1
No
More variation among various orders of
Salmonid-type 3
V
R 1
T
No
fishes than across other vertebrate
Bonytongue-type 1
F
R W
1
No
classes
Zebrafish-type 2
F
S Y
1
No
Exa
United States
Environmental Protection
Agency
-------
V'ER* Example SeqAPASS Level 3 - PPARy
United States
Environmental Protection
Agency
In silicomechanism for lack of Rosiglitazone binding to zebrafish PPARy is severe steric
hindrance from Gly312Ser and Cys313Tyr mutation
Comparing positions 312 and 313 of human to other species
Taxa
Species
Position
312
Position
313
Susceptibility
Prediction
Relative
Transactivation
Mammal
Human
G
C
Yes
1.0
Mammal
Mouse
G
C
Yes
1.2
Mammal
Rat
G
c
Yes
1.2
Amphibian
Clawed Frog
R
c
No
0.06
Fish
Sea Bream
S
c
No
<0.006
Fish
Plaice
S
c
No
<0.006
Fish
Zebrafish
S
Y
No
<0.006
10
-------
5-EPA —» . A .
Strategic Approach
United States
Environmental Protection
Agency
Similar types of analyses applied to
GR
PPARa
RXRb
Selected a group of species that should capture maximum
diversity in response for these four NRs (& genomes available)
• Human
• Xenopus laevis
• Rainbow trout
• Japanese medaka
• Zebrafish
-------
oE
United States
Environmental Protection
Agency
Attagene EcoTox-2 Factorial assay
#
Name
Species
Latin names
1
GR
human
Homo Sapiens
2
GR
african clawed frog
Xenopus laevis
3
GR
rainbow trout
Oncorhynchus mykiss
4
GR
japanese medaka
Oryzias latipes
5
GR
Zebrafish
Danio rerio
6
PPARa
human
Homo Sapiens
7
PPARa
african clawed frog
Xenopus laevis
8
PPARa
rainbow trout
Oncorhynchus mykiss
9
PPARa
japanese medaka
Oryzias latipes
10
PPARa
Zebrafish
Danio rerio
11
PPARg
human
Homo Sapiens
12
PPARg
african clawed frog
Xenopus laevis
13
PPARg
rainbow trout
Oncorhynchus mykiss
14
PPARg
japanese medaka
Oryzias latipes
15
PPARg
Zebrafish
Danio rerio
16
RXRb
human
Homo Sapiens
17
RXRb
african clawed frog
Xenopus laevis
18
RXRb
rainbow trout
Oncorhynchus mykiss
19
RXRb
japanese medaka
Oryzias latipes
20
RXRb
Zebrafish
Danio rerio
21
ERa
human
Homo Sapiens
22
ER1
Zebrafish
Danio rerio
23
ER1
african clawed frog
Xenopus laevis
AR
human
Homo Sapiens
\—X
AR
Zebrafish
Danio rerio
• 14 chemicals in concentration-response
• Surface water extracts
12
M-61
AR-ZF
RXRb-FR
GR-FR
RXRb-Hu
PPARa-RT
PPARg-JM
GR-JM
PPARg-FR
AR-Hu
PPARg-Hu
ERa-Hu
PPARg-ZF
PPARa-ZF
PPARg-RT
ER1-FR
GR-ZF
M-32
M-19
PPARa-Hu
GR-RT
PPARa-JM
RXRb-ZF
RXRb-RT
M-06
GR-Hu
PPARa-FR
RXRb-JM
ER1-ZF
-------
i5rSip,o,",ion Tgst Chgmica Is
Test Chemical
Target
Rosiglitazone maleate
PPARg
Tributyl phosphate
PPARg, PXR
Prednisone
GR, AR
Troglitazone
PPARg, PPARa
Zileuton
PPARg; ALOX5
Bexarotene
RxRb
Gemfibrozil
PPARa
Butachlor
GR, AR (env)
Triphenyl phosphate
PPARg (env)
Fenofibrate
PPARa
Dexamethasone NaP04GR
Triphenyltin chloride
RxR, RAR
PFOA
PPARa (env)
Potassium PFHxS
PPARa (env)
-------
5-EPA
Results - PPARy
United States
Environmental Protection
Agency
Rosiglitazone maleate
Troglitazone-
Triphenyltin chloride-
Triphenyl phosphate-
Tributyl phosphate-
Zileuton-
T
T
T
* ^ J>°
<§> V
l/ J
-------
oE
United States
Environmental Protection
Agency
Results - PPARa
Gemfibrozil -
Fenofibrate-
PFOA-
Potassium PFHxS-
Log ACC
4
>
-------
Results - RXRB
United States
Environmental Protection
Agency
Bexarotene-
Triphenyltin chloride-l
4*
v*°
&
&
O* o*
.«¦
<5-
Rainbow trout and zebrafish RXRb
were less sensitive to RXRb ligands
than the other species tested.
16
-------
«>EfB nQf | ilfc r~~ Q
Environmental Protection £_J I III — I ^
Agency I \ CI O U I I O VJ I \
Predicted Susceptible
Taxa
Homo
sapiens
Xenopus
laevis
Oryzias
latipes
Oncorhynchus
mykiss
Danio rerio
Dexamethasone
Y
Y
N
N
Y
Dexamethasone NaP04
Prednisone
Butachlor
t——I——I——I——r
Log ACC • Predictions were qualitatively accurate
4 for dexamethasone but reflected
l_gss
potent different sensitivity, not overall
susceptibility
o
More # Need t0 nnetabolically activate
potent prednisone to the GR-active prednisolone
complicates interpretation
17
-------
oE
United States
Environmental Protection
Agency
Application to Environmental Monitoring
_0J
Q.
E
03
LO
03
<£> * <£? ^ oD'
* o
-------
AEPA ^ . .
sr1Conclusion
• Effects-based monitoring employing human cell lines using human nuclear receptors
(hNR) are likely to yield different conclusions than if fish NRs were employed (at least
for PPARy, PPARa, RXRp, and GR).
• Variations among different orders offish may be as substantial as across other classes
of vertebrates.
• Different chemical-specific profiles across species were consistent with a previous
assumption that level 3 SeqAPASS analyses based on specific ligand-chemical
interactions may not apply universally across relevant chemical space.
• Complicates the ability to select a minimum number of species to capture
maximum variability in sensitivity.
Screening of additional chemicals using the XS-2 Factorial Assay may yield new insights
that improve the ability to predict cross-species susceptibility based on aa sequence.
-------
AEPA
United States
Environmental Protection
Agency
References
• Cavallin et al. Effects-Based Monitoring of Bioactive Chemicals Discharged to the Colorado River
before and after a Municipal Wastewater Treatment Plant Replacement Environmental Science &
Technology 2021, 55, 2, 974-984. httos://doi.org/10.1021/acs.est 0c05269
• Lalone et al. Sequence Alignment to Predict Across Species Susceptibility (SeqAPASS): A Web-
Based Tool for Addressing the Challenges of Cross-Species Extrapolation of Chemical Toxicity.
Toxicological Sciences, Volume 153, Issue 2, October 2016, Pages 228-245,
httos://doi. org/10.1093/toxsci/kfw119
• Medvedev et al. Harmonized cross-species assessment of endocrine and metabolic disruptors by
EcoTox FACTORIAL assay. Environmental Science & Technology 2020 54, 19, 12142-12153.
https://doi. org/10.1021/acs. est. 0c03375
-------
oEPA
United States
Environmental Protection
Agency
High Throughput Transcriptomics:
A Multi-Species Approach
Presented by Kevin Flynn
to
US EPA BOSC
Chemical Safety Subcommittee Meeting
Office of Research and Development
Center for Computational Toxicology and Exposure
-------
AEPA
United States
Environmental Protection
Agency
ORD Strategic Research Action Plan
CSS. 1.7:
Develop, evaluate, and apply non-mammalian high-throughput toxicity tests for
priority endpoints and pathways in ecological species for ecological risk assessment
CSS.4.4:
Develop rationale and case studies that apply AOPs and HTT data to inform test-order
decisions and establish scientific support for waiving testing requirements for pesticides
Office of Research and Development
Center for Computational Toxicology and Exposure
-------
oE
United States
Environmental Protection
Agency
Program Office Feedback & Support
Program Office support
• Office of Pesticide Programs (OPP)
• Office of Pollution Prevention and Toxics (OPPT)
EPA/ORD Partner Needs and Use
OCSPP; OW:
Ecological risk assessments address species across diverse taxonomic groups, many of which have limited or no available
data. The clear majority of HTT methods are based on either human or mammalian in vitro systems, which results in an
under-representation of pathways that are relevant and perhaps unique to non-mammalian taxa
OPP:
"The proposed research is responsive to EPA's commitment to reduce its reliance on resource intensive whole animal
testing and to better allocate limited resources to where they are most needed through the use of more effective
screening tools"
• evaluate taxonomic domain of applicability of HTT NAMs
• determine relevance to regulatory assessment endpoints
• provide evidence that risk estimates across taxa are suitable
• critical effort to turning theory into practice (risk estimates)
OPPT:
"HTT methods are required for OPPT to rapidly and efficiently screen and prioritize new and existing chemical substances
under the Toxic Substances Control Act (TSCA)."
• meet scientific standards under section 26(h) of TSCA
• reduce testing on vertebrates under section 4(h) of TSCA
• determine unreasonable risk under sections 5 & 6 of TSCA
Demonstrating quantitative relationships between
transcriptomic responses and phenotypic alterations
in apical assessment endpoints across multiple taxa
will provide a compelling means of promoting the
adoption of such methods and incorporation of such
data in ecological risk assessments - OPP comment
to Eco-trans Output
The effort should provide an understanding of
how the battery of assays relate an AOP where a
molecular initiating event leads to a sequence of key
events that culminate in an adverse outcome at the
whole organism level and how such methods can be
used to effectively screen for effects across a wide
taxonomic and chemical space. OPP has a wide array
of in vivo data on pesticides and these data could
serve as a means of assessing the extent to which the
battery of assays is predictive across chemicals/taxa -
OPP comment to Eco-trans Output
OPP has a broad array of in vivo data across multiple
taxa with which to compare transcriptomic responses
once suitable linkages have been demonstrated. OPP
is interested in working with ORD to identify
"model" chemicals - OPP comment to Eco-trans
Output
-------
&EPA
United States
Environmental Protection
Agency
A Chemical Numbers Problem
U.S. EPA Strategic Plan (2018-2022), Objective 1.4,
Ensure Safety of Chemicals in the Marketplace
Problem Statement:
Tens of thousands of chemicals
hundreds more are introduced
Only a small fraction has been
potential risks to human health
"Too many chemicals, too little data"
Office of Research and Development
Center for Computational Toxicology and Exposure
-------
£EPA
United States
Environmental Protection
Agency
A Biological Numbers Problem
Gene Coverage
ToxCast
Not in
ToxCast
"Throughout the development and execution of ToxCast and Tox21, key
limitations of the current suite of HTS assays have been identified (Tice, et
a!., 2013). The limitations include inadequate coverage of biological
targets and pathways" Thomas et al. 2019
The Eco Data Gap:
Humans are just a tiny fraction of the biological diversity we
are charged to protect.
Many genes/pathways are conserved
Unique physiology in other kingdoms, phyla, classes...
Nematodes
Chordates
Crustaceans
Arachnids
Other
Plants
Moliusks
Insects
Office of Research and Development
Center for Computational Toxicology and Exposure
-------
&EPA
United States
Environmental Protection
Agency
HTP Eco Assay DeveDopment
Daphnia magna Pimephales promelas Chironomus dilutus Raphidocel/s subcapitata
• Modify standard protocols and methods to allow rapid toxicity tests with small aquatic organisms in
96-well plates - 4 species
• Conduct exposures with diverse chemicals (ex. metals, neonics, pharmaceuticals, PFAS)
• Compare traditionally derived LC50 values to LC50 values calculated from 96-well plate-based
exposures
• Use RNA-seq data to calculate transcriptomic-based point-of-departure (tPODs) that can be
anchored to apical responses
Office of Research and Development
Center for Computational Toxicology and Exposure
-------
SERA
United States
Environmental Protection
Agency
HTP Eco Assay DeveOopment
Control
Replicates /
v
A
B
C
D
E
F
G
H
V
oooooooooooo
24 h exposure
8 9 10 11 12
Phenotypic anchoring
• survival
behavior
growth?
Office of Research and Development
Center for Computational Toxicology and Exposure
Species
Guideline Test Method
Age at Start
Temp
Daphnia magna
850.1010 Aquatic Invert Acute Toxicity
72-hour
20° C
Pimephales promelas
850.1075 Fish Acute Toxicity
24-hour
25° C
Chironomus dilutus
850.1790 Chironomid Sediment Toxicity
3rd instar
20° C
Raphidoceiis subcapitata
850.4500 Aigai Toxicity
Log-phase
24° C
Exposures Design
1 ml deep 96-well plates
12 concentration - 8 replicates per concentration
1 individual per well (algae ~5 x 104 cells/ml)
24-hour static exposures
phenotypic endpoints assessed
animals: survival and behavior
algae: cell viability & division, photopigments
then after homogenization, RNA extracted for transcriptomics
Species
Time to Load Plate
Control 24-hour Survival
RNA Qty per Well
Daphnia magna
~45 minutes
72-hour
~1000 ng
Pimephales promelas
~3Q minutes
24-hour
~1500 ng
Chironomus dilutus
~60 minutes
3rd instar
~900 ng
Raphidoceiis subcapitata
~10 minutes
Log-phase
~300 ng
-------
SERA
United States
Environmental Protection
Agency
HTP Eco Assay DeveOopment
Toxicant
Chemical
initiator
LC50s:
Published vs 96-well Exposure
.1 1 10
Plate LC50
Office of Research and Development
Center for Computational Toxicology and Exposure
Macro-Molecular
Interactions
Cellular
Responses
Organ
Responses
TndivicIuaF
Responses
Increasing levels of biological organization
Population
Response
Linking to apical endpoints essential
Apical Endpoints
• Survival &
• Reproduction
• Growth & orc?
Behavior
"Imageable" measurements
-------
SERA
United States
Environmental Protection
Agency
Eco Transcriptomics Applicability
NTP
National Toxicology Program
U S. Dopartmont of Health and Human Services
NTP Research Report on
National Toxicology
Program Approach to
Genomic Dose-Response
Modeling
NTP RR 5
APRIL 2018
A.3 Global Comparison of POD and BEPOD
6
l-ioo
1 10 100 1000
Apical Potency Value with 1/2 log range
Figure 1-1. Comparison of the Most Sensitive Apical Vz Log Potency Range to the Most Sensitive GO
Biological Processes BEPOD
Data from Figure 1-Figore 13 in tins document were compiled to allow a lareer scale corrpprisoa of apical and gems set-based
biological potency estimates. The most sensitive 3pical potency values (NOAEL or BMD) frcoi guideline imorify assessments
are plotted on tie x-asis and tie BEPOD range (BMDi.-BMD-BMD-j) from die GO Biological Processes analysis from 4- or
5-day GDHS studies are plotted on the y-axis. A diagonal 1-to-l line is drawn as reference to perfect agreeivnt between tie
potency values. The points to The left of tie line demonstrate mote sensitive apical endpoims, niereas tiose to the right exhibited
more sensitive BEPODs. Overall, the apical and BEPOD \-alues strongly agree, as indicated by R: = 0.89.
Number of mammalian studies have shown short-term
transcriptomics-based PODs are predictive of apical potency.
Generally, within /2 log.
Office of Research and Development
Center for Computational Toxicology and Exposure
ELSEVIER
Toxicology and Applied Pharmacology 378 (2019) 114634
Contents lists available at ScienceDirect
Toxicology and Applied Pharmacology
journal homepage: www.elsevier.com/locate/taap
Transcriptomic points-of-departure from short-term exposure studies are
protective of chronic effects for fish exposed to estrogenic chemicals
Florence Page-Lariviere, Doug Crump, Jason M. O'Brien
Naaonai WltSife Research Center, Envimnmem and CBmiSe Change Canada. Ontario. Canada
H)
F. Pagf-Larmire. a at
A
Toxicology and Applied Pharmacology 378 (2019) 114634
/'
-------
SERA
United States
Environmental Protection
Agency
HTP Eco Transcriptomcs
Challenge.gov
Govern men! Challenges, Your Solutions
EcoTox TARGET Challenge
Develop high quality, low-cost tools that assess global gene expression in
common aquatic toxicity test organisms
S3 TOTAL CASH PRIZES OFFERED: $300,000
=TYPE OF CHALLENGE: Scientific
2t PARTNER AGENCIES | FEDERAL: U.S. Army Engineer Research and Development Center
12 PARTNER AGENCIES | NON-FEDERAL: DoW, Environment and Climate Change Canada, European Commission
Joint Research Centre, Syngenta
EB SUBMISSION START: 03/19/202012:00 PM ET
O SUBMISSION END: 06/15/202111:59 PM ET
Detection/analysis
technology
TARGET
EcoTox Challenge
\
&EPA
Innovation: $300,000 EPA Research Challenge Prize Available
EPA is looking for innovators who can help usher in the
next generation of Technology Advancing Rapid
Gene Expression-based Testing (TARGET).
The Agency is offering a $300,000 prize to the
company, organization, or team that can provide high
quality, low cost, technologies/platforms for evaluating
global gene expression in samples (RNA or tissue
homogenates) from four commonly used aquatic
toxicity test organisms:
* Pimephales promelas (a fish)
¦ Daphnia magna (a crustacean)
* Chironomous dilutus (an insect; formerly Chironomous tentans)
¦ and Raphidocelis subcapital (a green algae)
Think you have a winning technology? Learn more at:
Office of Research and Development
Center for Computational Toxicology and Exposure
Commercial
Development
4 "Solvers"
Commercially available
Low cost (<$50/sample)
High quality
Maxima! coverage
TARGET
I
EcoTox Challenge
-------
SERA
United States
Environmental Protection
Agency
Eco Transcriptomocs Data Analysis
Transcriptomics Analysis Workflow
• not re-inventing the wheel
• mirror ToxCast data analysis
BMDExpress2
Chemical +
Eco-species
Eco-
Transcnptomics
Raw data
processing
Data QC
Differential
expression
analysis
Pathway
Aggregation
Concentration Response
'• Si:
- h.#
¦i-M
¦ •>'sSV
¦¦M
te-.
Oown
Not S«)
c5H ¦
:
IS
i
* . L
x
_
34,6 H
« B
i
44 I
I E5.0
Concentration Response
Concentaaon
~ hXltDAi—MhH VIM !>L0Ta
Office of Research and Development
Center for Computational Toxicology and Exposure
symmetric, uriimodal
mud.
uniform bi modal multimodal
BMD Grouping
All BMDs - median
X lowest (ex. 20)
Percentile (ex. 10th)
1st mode
All BMDs from lowest GO terms or REACTOME pathways
Median BMDs from 20 lowest GO terms or REACTOME pathways
-------
&EPA
United States
Environmental Protection
Agency
Current Status
Mcc6lerating tno ^dOG of C/h6rniC3l Risk Asssssrn0nt
APCRA '.»•
^ ^" • 4*^ ~ • j6.
Derive transcriptomics-based points of departure for 20 chemicals
Testing with fathead minnow only
Compare with traditional apical PODs
Evaluate hypothesis that tPODs are protective relative to apical
Includes chemicals of direct interest to Program Offices and partners
Workflow in Brief
RNA-seq data was obtained from each well; all raw reads were assembled into transcript models,
aligned with annotations, counted, normalized, and log2 transformed for each transcript
• Low count feature filtering: any given feature had to have a count of 10 or more in a minimum of
4 samples or that feature was filtered out
• Differentially expressed genes (DEGs) determined by NTP guidelines and transcriptomic POD
for a chemical defined as median or 10th percentile POD of all DEGs
(https://ntp.niehs.nih.gov/publications/reports/rr/rr05/index.html)
Office of Research and Development
Center for Computational Toxicology and Exposure
-------
AEPA
United States
Environmental Protection
Agency
HTP Eco Assay Development
Chemicals
Chemical Class
Rationale
Data Product Support
CuS04, IMiS04, ZnS04
metal
OW; ease of exp.; mouse & RBT data
APCRA case study; 4 eco-species
Clothianidin, Thiacloprid, Imidacloprid
Neonicotinoid
OPP
APCRA case study; 4 eco-species; Challenge
Flupyradifurone
Butenolide
OPP
APCRA case study; 4 eco-species
Sertraline, Fluoxetine, Paroxetine
SSRI
Existing data at GLTED
APCRA case study; 4 eco-species
Atrazine, simazine, cyanazine
Herbicide
Herbicide
Challenge; 4 eco-species
Methoxyfenozide, tebufenozide,
methoxyfenozide
Carbohydrazide
Insecticide
Challenge; 4 eco-species
Parathion, methidathion, fenthion
Organophosphate
mouse data
APCRA case study; 4 eco-species
dibutyl, diethylhexyl, butylbenzyl
phthalates
Phthalates
TSCA high priority
APCRA case study; 4 eco-species
~20 specific PFAS
PFAS
PFAS plus up; small # in vivo
4 eco-species
50 -100 additional
diverse
St RAP
4 eco-species
Office of Research and Development
Center for Computational Toxicology and Exposure
-------
SERA
United States
Environmental Protection
Agency
BMDiHi;
Express2J
Draft Data
A»NTP AEPA
[Sensitive apical
[tPOD] < endpoirit]
it
[tPOD]
[Sensitive apica
«<
endpoint]
¥
Chemical
Transcriptomic POD
(10th and median)
96-hour
LC50
Mortality-based
POD
CuS04
0.005 mg/L
0.03 mg/L
0.3 mg/L
0.2 mg/L
ZnS04
0.063 ug/L
0.23 ug/L
2.2 mg/L
3.2 mg/L
NiS04
0.146 mg/L
0.33 mg/L
6.2 mg/L
3.9 mg/L
Imidacloprid
0.03 ug/L
8.8 mg/L
173 mg/L
> 10 mg/L
Flupyradifurone
0.0226 ug/L
1.3 mg/L
Not in ECOTOX
> 10 mg/L
Clothianidin
0.2 ug/L
8.1 mg/L
0.5(104) mg/L
> 10 mg/L
Thiacloprid
0.036 mg/L
57.2 mg/L
104 mg/L
85 mg/L
Sertraline 0.001 mg/L 0.6 mg/L 0.1 mg/L 0.9 mg/L
Fluoxetine 0.003 |ig/L 0.02 mg/L 0.2 mg/L 0.8 mg/L
Paroxetine 0.002 mg/L 1.0 mg/L 3.5 mg/L 1.1 mg/L
Office of Research and Development
Center for Computational Toxicology and Exposure
-------
SERA
United States
Environmental Protection
Agency
Upcoming Work
Assay Development
Verify water quality parameters
• dissolved oxygen
• pH
• ammonia
Chemical bioava lability
POD Calculation for CuSo4 in each Volume
BMDExpress2 Results
Volume Format
CUP
24WP
96WP
#DEGs passing NTP filters1
128/369
52/159
108/208
Median POD (mg/L)
0.0445
0.045201
0.025
CUP vs 96WP
CUP vs 24WP
24WP vs 96WP
Office of Research and Development
Center for Computational Toxicology and Exposure
Overlap of DEGs
Validation
Transcriptomics
• Complete Challenge
• platform development
• genome annotation
• Definition/Implementation of analysis pipeline
• Assess variability focused on tPODs
• intra/inter exposure plate
• between exposure plates
• appropriate replication
CuS04 — Replicate Sub sampling 12x: tPODs
0.015-
D-OIO-
_e
Q
o
9b
O-005-
O-OOO-
4 5 6
Number of Replicates
-------
&EPA
United States
Environmental Protection
Agency
Contributors
The "We"
ORD CCTE GLTED-MIB: Adam Biales, David Bencic, Robert Flick, John Martinson
ORD CCTE GLTED-STB: Kevin Flynn, Dan Villeneuve, Kathy Jensen, Jenna Cavallin
ORD CCTE GLTED-TTB: Russ Hockett, Teresa Norberg-King
ORD CCTE BCTD-RADB: Josh Harrill
ORD CCTE BCTD-CTBB: Logan Everett
ORISE FELLOWS: Michelle Le, Kelvin Santana-Rodriguez, Kendra Bush, Monique Hazemi
Acknowledgements
Christina Inglis, John Prindiville, Andy Nong, Jason O'Brien, Florence Page-Lariviere
I ^ | Environment and
Climate Change Canada
Environnement et
Changement climatique Canada
Office of Research and Development
Center for Computational Toxicology and Exposure
-------
Rpfprpnrpq
Environmental Protection J m I I I I
• U.S. EPA Strategic Plan (2018-2022), Objective 1.4, Ensure Safety of Chemicals in the Marketplace
• Thomas, R. S., Bahadori, T., Buckley, T. J., Cowden, J., Deisenroth, C., Dionisio, K. L.,... & Williams, A. J. (2019). The
next generation blueprint of computational toxicology at the US Environmental Protection Agency. Toxicological
Sciences, 169(2), 317-332.
• Page-Lariviere, F., Crump, D., O'Brien, J. M. (2019). Transcriptomic points-of-departure from short-term exposure
studies are protective of chronic effects for fish exposed to estrogenic chemicals. Toxicology and Applied
Pharmacology, 378, 114634.
• EcoTox TARGET Challenge: https://www.challenge.gov/challenge/ecotox-challenge/
• National Toxicology Program Approach to Genomic Dose-Response Modeling:
https://ntp.niehs.nih.gov/publications/reports/rr/rr05/index.html
Office of Research and Development
Center for Computational Toxicology and Exposure
------- |