=Paper= {{Paper |id=Vol-2931/ICBO_2019_paper_11 |storemode=property |title=TXPO: A Toxic Process Ontology for Better Understanding of Drug-Induced Liver Injury |pdfUrl=https://ceur-ws.org/Vol-2931/ICBO_2019_paper_11.pdf |volume=Vol-2931 |authors=Yuki Yamagata,Yoshinobu Igarashi,Noriyuki Nakatsu,Hiroshi Yamada |dblpUrl=https://dblp.org/rec/conf/icbo/YamagataINY19 }} ==TXPO: A Toxic Process Ontology for Better Understanding of Drug-Induced Liver Injury== https://ceur-ws.org/Vol-2931/ICBO_2019_paper_11.pdf

TXPO: A toxic process ontology for better understanding of drug-induced liver injury

Yuki Yamagataa, Yoshinobu Igarashia, Noriyuki Nakatsua, Hiroshi Yamadaa
a
Laboratory of Toxicogenomics Informatics, National Institutes of Biomedical Innovation, Health and Nutrition,
Ibaraki, Osaka, Japan

Abstract system (TOXPILOT). Here, we discuss the current state of our
work.
Elucidating the mechanism of toxicity is crucial in drug safety
evaluations. We focus on toxic processes and developed a toxic
process ontology, designated TXPO. Here, we outline the TXPO, Methods
which systematizes toxic processes within the liver in a con-
sistent manner. The TXPO makes processes explicit across
TXPO development
granularity using a functional decomposition tree. Concerning
the course of toxic processes, we present a framework of causal From textbooks [2-6] we researched drug-induced hepatotoxic
relationships between processes from latent to toxicity manifes- mechanisms and obtained information about toxic courses and
tation. In applied work, we introduce a prototype of TOXPILOT, related processes, molecules and their roles, and biological
a toxic process interpretable knowledge system. TOXPILOT structures. Next, we searched for the latest information from
provides visualization maps of the toxic course, which facilitates toxic course-related articles using PubMed search terms in Table
capturing the comprehensive picture for understanding toxicity 1.
mechanisms. Our ontological approach will help develop new Table 1 PubMed search terms for hepatotoxic course
knowledge regarding drug safety evaluations.
Keywords:
ontology; process; drug-induced liver toxicity

Introduction

Drug-induced liver injury is a major cause of drug withdrawal
from the market and discontinuation of drug development [1].
Therefore, safety assessments during the early stages of drug
development are required. Toxicology is a scientific discipline
that examines the biological effects (toxic effects) of substances
such as chemical compounds, drugs, and drug candidates. We
developed a hepatotoxicity prediction informatics system with
the aim of developing safety biomarkers during the early stages
of drug development. We conduct hepatotoxicity predictions
based on computational approaches using toxicogenomics data We used the ontology editing tool Protégé 5.2.0 [7] to develop
and machine learning. In order to promote data-driven research the TXPO in the Web Ontology Language (OWL) and HermiT
and appropriately assess safety, it is necessary to explain compu- reasoner [8] as a Protégé Plug-in.
tational predicted results in light of the relevant mechanisms. Figure 1 shows examples of the TXPO development process.
However, the mechanisms of hepatotoxicity are complex, in part First, 1) each toxic course was defined, and related information
because the liver is the site of drug metabolism, the results of was annotated using the Annotation Properties. Next, 2) the pro-
which can affect a wide variety of biological structures and cesses constituting each toxic course were described using a 'has
functions. For safety management, it is desirable to systematize part' relation as Object Property. Then, 3) each process was gen-
the necessary knowledge from a consistent viewpoint. eralized using an is-a hierarchy: processes common to multiple
toxic courses, biological processes, and biomedical-independent
To better clarify toxicity mechanisms, in the present study, we processes. Furthermore, 4) each process was decomposed into
developed a toxic process ontology (TXPO). The TXPO system- subprocesses (has part relation), and 5) the biological structure
atizes a wide variety of toxicological terms involving hepatotox- in which the process takes place was described (occurs in). In
icity processes. We also modeled a representation framework addition, 6) molecules/drugs and their roles in the process were
that appropriately describes toxic courses. In applied work, we also defined. Finally, 7) causal relationships between process
developed a prototype toxic process interpretable knowledge were defined by using a 'has result' relation.

Copyright © 2019 for this paper by its authors. Use permitted under Creative Commons License Attribution 4.0 International (CC BY 4.0).
system. The TXPO file is stored in the ontology library, and
(1) the file is converted to RDF format represented by a triple
Subject, Predicate, and Object by Protégé. The RDF data
are then stored in an RDF triple store using Apache Fuseki
[19] to construct the SPARQL endpoint. Regarding the web
(2) application system for TOXPILOT, necessary information
is dynamically acquired via SPARQL queries. Moreover,
(3) TOXPILOT generates graphs using D3.js [20] of the Ja-
vaScript library.

Results

(4) Development of the TXPO

Outline of the TXPO
(6) (5)
Figure 2 shows an overview of the TXPO, which is a
three-layer model organized in an is-a hierarchy of general
terms to specialized toxicologic terms. The top layer is do-
main-independent (domain-neutral) and provides general
(7) terms. Most of the entities in the top layer refer to the upper
Figure 1 Examples of TXPO development ontology, Basic Formal Ontology (BFO). Upper ontologies
support generic categories and relations based on a philo-
In generalizing the is-a tree construction, we reused existing sophical orientation. Accordingly, we could construct our
ontologies. Domain- independent general entities were based on ontology with inheritance of the intrinsic nature in a consistent
BFO [9], and biomedical entities were imported manually from manner. All entities of the TXPO are classified into the basic
existing ontologies in NCBO BioPortal [10]. These biomedical categories continuant or occurrent. Continuant refers to an enti-
ontologies include UBERON [11], Cell Ontology [12], NCBI ty that persists, endures, or continues to exist through time while
Taxon [13], ChEBI [14], Gene Ontology [15], PATO [16], maintaining its identity and includes objects, roles, and qualities.
INOH [17], and Ontology of Genes and Genomes (OGG) [18]. An object is an independent continuant, such as a thing. Roles
and qualities are dependent continuants that can only exist de-
TOXPILOT development
pending on something else. Occurrent includes entities that un-
TOXPILOT consists of an ontology library, a Resource De- fold over time, such as processes.
scription Framework (RDF) database, and a Web application
The intermediate layer is biomedical domain dependent and

Figure 2. Overview of the ToXic Process Ontology (TXPO). The TXPO contains an is-a hierarchy that is organized into three
layers: the top layer contains general terms, mostly derived from the Basic Formal Ontology. The intermediate layer contains
biomedical terms, and the lower layer contains more granular, toxicology terms.
consists of entities commonly used in biomedicine. As lower (1) Functioning Process
entities of continuant, biological structures such as molecules, Many biological defense processes function to protect organ-
compounds, organelles, cells, organs, and species are defined. isms from toxicity-associated injury. Therefore, we focused on
The open community OBO Foundry [21] seeks to share functioning processes in the present study. Functioning process-
knowledge and standardize terms among the biological commu- es in organisms are diversified in granularity from the molecular
nity, and OBO ontologies utilize BFO as an upper ontology. level to the organelle, cell, tissue, and organ level. In order to
Accordingly, the TXPO imports existing terms and reuses them define functioning processes in a consistent fashion, we sys-
from biomedical ontologies of OBO foundries. These terms in- tematize the functioning tree based on functional ontology [22,
clude anatomic structures from UBERON, cells from Cell On- 23]. As an ontological engineering approach, functional ontolo-
tology, organisms from NCBI Taxon, compounds from ChEBI, gy defines general functions based on changes in the state of the
biological processes and cellular components from Gene Ontol- input-output relationship between physical things and models
ogy, qualities from PATO, some molecule families from INOH, the functional knowledge (Fig. 3 (a)). As a basic idea, a func-
and genes from OGG. tioning process can be categorized into receiving, making exist-
The lower layer encompasses entities specific to toxicology ent, and generating groupings based on the number of focused
(i.e., entities that are toxicological domain dependent). inputs and outputs of the target. The making existent category
can be further subdivided into changing an operand and chang-
Process in the TXPO ing relationship between operands classifications. Changing an
Process is a central category in the TXPO. In order to elucidate operand includes changing qualities such as concentration, pres-
a toxicity mechanism adequately, we provide two sub- sure, volume, etc. Examples of subtypes of changing relation-
categories, namely, primitive process and process sequence. The ship between operands are transmitting and separating. Sub-
former is defined as a single unit of process, whereas the latter is types of separating include decomposing, splitting, and detach-
defined as a series of processes, which includes pathways and ing. Based on these terms and by specializing their use, we de-
toxic courses. veloped the functioning is-a hierarchy in the TXPO (Fig. 3 (b)).
The intermediate layer is biomedical domain dependent. For
instance, a lower level of transmitting includes biological

(a) Examples of functions in functional ontology

Figure 3 functioning process
(b) The TXPO functioning is-a hierarchy
transport processes such as nuclear transport and Golgi vesicle functioning processes of the cellular system parts (i.e., orga-
transport. Decomposing includes proteolysis and lipid degrada- nelles).
tion; splitting includes cell division; and detaching includes
complex dissociation. These processes are generally consistent (3) Toxic course
with the GO Biological Process, as some of the GO biological In toxicology research, elucidating the mechanism of toxicity is
processes can be interpreted as functional processes common to crucial for safety management. Toxicity mechanisms are general-
biomedicine. ly explained in terms of multiple processes, such as toxicant de-
livery, biological defense processes, cellular dysfunc-
The lower layer is a toxicology domain dependent. In this tion/dysregulation, and cell death. Therefore, in the present study,
study, we define a "toxic process" as a process that constitutes a we focused on toxic courses. As a subtype of the process se-
specific toxic course. For example, by specializing the biomedi- quence, the TXPO defines a toxic course as a series of processes
cal process "apoptotic process (GO:0006915)", we define an in an organism from latency to the manifestation of toxicity,
"apoptotic process [ER stress]" that constitutes a course of ER which is not part of the normal life of the organism. Subtypes of
stress, and an "apoptotic process [Phospholipidosis]" that consti- the toxic course include specific themes, such as ER stress, gluta-
tutes a course of Phospholipidosis, and so on. thione depletion, phospholipidosis, lipidosis/fatty liver, ground
glass appearance of hepatocytes, and eosinophilic granular de-
One of the difficulties of capturing a toxic process is that generation.
some toxic effects are protective responses to xenobiotic
substances (drugs) [24]; hence, to understand the toxicity mech- In the present study, we developed a framework called the “tox-
anisms appropriately, we also regard a process functioning as a ic course map” to represent toxic courses uniformly. The map
biological defense in the specific toxic course as a "toxic pro- represents a toxic course as causal relationships between process-
cess." es (Fig. 4). With regard to development of toxicity, we applied
the imbalance theory [26]. In the present study, supply indicates
Developing a toxicity-dependent process subtree is based on the functioning processes associated with biological defense and
the low-hanging fruit policy. From toxicology-related textbooks maintaining homeostasis, and demand refers to toxic activity. As
and published articles, terms were extracted and manually anno- illustrated in Figure 4, in the imbalance model, the basic units are
tated. as follows:
Here, as a function-related process, in addition to the function- 1) a functioning process (supply) for biological defense and
execution process, the TXPO defines meta-functioning process- maintaining homeostasis;
es. Meta-functioning processes are functioning processes specif- 2) a functional demand process (demand) as toxic activity;
ic to other functions and include controlling, for example. Sub-
types of controlling include the regulation of apoptosis and cell 3) balance/imbalance between toxic activity and defense pro-
cycle control. cesses; and
4) outcome from organelles, cells, or tissues to the organ ex-
(2) Decomposition of Functioning
hibiting toxicity manifestations
The TXPO specifies a functioning process based on a func-
tion decomposition framework. As an ontological engineering
approach, a device (system) consists of sub-devices (sub-
systems). In a function decomposition tree, the whole function of
a system is achieved by a sequence of sub-functions of the sub-
systems. As biological functions can be considered specializa-
tions of systemic functions [25], in the present study, we attempt-
ed to clarify the functioning process of biological structures for
each granularity based on the whole-part relationship (part of/ has
part relationship). At the cell level, we regard a cell as the system
and cell components such as organelles as system parts. Figure 3
(c) shows an example describing how the cell system functions
from a decomposition perspective. In the toxic course of endo-
plasmic reticulum (ER) stress, for example, the accumulation of
drugs such as tunicamycin in the ER is known to initiate protein Figure 4 Representation framework of a toxic course
unfolding. Therefore, the cell system executes the "protein quali-
ty control" function as a biological defense function. Here, we The degree of functioning performance can change according
can say that the cell system consists of subparts: the ER, ribo- to changes in demand; however, if demand exceeds the perfor-
somes, nucleus, and cytoplasm. During the early stages of ER mance of functioning, an imbalance occurs and results in an
stress, the sub-functioning process "protein refolding" is carried outcome that is no longer latent and manifests toxicity. Table 2
out in the ER. The ER receives input regarding an unfolded pro- shows examples describing the imbalance framework in ER
tein, and after executing the refolding function, the ER output stress.
consists of the refolded protein. In addition, "translation attenua-
tion" is also carried out by the ribosomes to suppress production
of new proteins, which supports protein refolding. However, if
the refolding process is not sufficient, then, “regulating gene ex-
pression” can occur in the nucleus, and in the cytoplasm, "protein
degradation" is executed, with the unfolded protein serving as the
input and its degradation product as the output. Thus, the cell
system achieves protein quality control through specific sub-
(5) Relationship between entities
As of February 1, 2018, the TXPO defined approximately 6000
Table 2. Examples of imbalance in ER stress entities, and Figure 5 shows the major relationships between
Toxic action Functioning
Granularity (Demand) Imbalance (Supply) Outcome
terms defined in the TXPO.
Accumulation
ER Unfolding > Refolding of unfolded
protein Applications
Producing Degrading
Protein aggre- We developed a prototype toxic process interpretable support
Cytoplasm unfolded > unfolded
gate formation
protein protein knowledge system, known as TOXPILOT. The TOXPILOT pro-
Autophagy vides varied useful information based on the TXPO (Fig. 6). The
Increasing
Cell (removing Abnormal ER
(hepatocyte)
protein aggre- >
protein formation TOXPILOT visualizes toxic course maps (Fig. 6 (a)), as de-
gates scribed in the previous section. Since our map can visualize mol-
aggregates)
Apoptosis
Accumulation ecules that participate in toxic processes, we can apply the map to
Increasing (removing facilitate explanation of biomarkers for toxicity prediction by
Tissue > of abnormal
abnormal cells abnormal
cells)
cells machine learning. Our preliminary data show that by using maps,
Cell death in vivo and in vitro data of predicted marker genes of liver toxici-
Organ (increasing
>> Cell survival Liver failure ty can be comparatively analyzed. As a result, we can identify
(liver) apoptosis,
necrosis)
genes predicted to participate in common processes in ER stress
Cell death based on rat in vivo and human in vitro analyses. Thus, this toxic
Organ (increasing
<<
Cell prolif- Liver carcino- course map facilitates evaluation and extrapolation to humans for
(liver) apoptosis, eration genesis translational research.
necrosis)
The TXPO also provides process maps (Fig. 6 (b)), in which
(4) Role
sub-processes can be displayed according to the whole-part rela-
In general, a molecule plays multiple roles in the body. There- tionship of systemic functioning across granularities. These maps
fore, in the present research, we tried to explicate the roles of also enable visualization of pathologic findings associated with a
molecules participating in specific processes in the toxic course. process.
For example, in ER stress, GRP78 participates in the protein re-
folding process and can play the role of a "chaperone" that assists The TXPO also provides a general course map that visualizes
protein refolding (Fig. 5). GRP78 also plays the role of "autopha- general toxic courses common to multiple specific toxic courses
gy inducer" in the positive regulation of autophagy process dur- (Fig. 6 (c)). In safety evaluation, toxicologists sometimes want to
ing ER stress. As viewed relative to the role of a molecule, the know whether one phenomenon that occurs in a particular toxic
TXPO contributes to identifying biomarkers that participate in course could occur in other toxic courses. For instance, in the
the turning points of processes that cause cell injury during the course of lipidosis, "lipid accumulation" can cause "increasing
course of toxicity manifestation. As for drugs, TXPO makes ex- hepatocyte volume." The TXPO system extracts information
plicit the role of drugs in a specific toxic process. For example, from the RDF database by SPARQL and automatically generates
tunicamycin plays a 'protein glycosylation inhibitor' role and par- a general course map. In the general course map, common pro-
ticipates in the negative regulation of glycosylation process. Tu- cesses are represented as large nodes. As a result, users can ob-
nicamycin also plays an 'apoptosis inducer' role in the positive tain information indicating that "increasing hepatocyte volume"
regulation of apoptosis in the liver (Fig. 5). is common to other toxic courses, such as cholestasis. Moreover,
users can obtain information regarding different causes associat-
ed with other courses. Each toxic course is colored, so users can
see easily that, for example, "bile acid accumulation" occurs spe-
cifically in the course of cholestasis.
Our system also provides a function for searching routes from
specific processes (Fig. 6 (d)). When users want to conduct retro-
spective analyses, TOXPILOT provides an illustration of ‘up-
stream’ of the focused process in the toxic course, which can help
identify critical causes during the early stages of toxicity. In the
same way, if users wish to know how a process unfolds with the
progress of the toxicity development, our system provides a
‘downstream’ illustration that supports severe manifestation risk
management.

Figure 5 Examples of TXPO relationships
(a) Toxic course map visualizes a toxic course as causal rela-
(b) Process map visualizes the selected process. Systemic functioning decom-
tionships between processes and visualizes molecules in-
position allows sub-processes to be displayed across granularities.
volved in toxic processes. The left window shows the is-a
hierarchy, and highlight shows the currently selected toxic
course. The lower window shows information based on the
TXPO.

(d) Route search provides the upstream
or downstream paths of the selected pro- (c) General course map visualizes general toxic courses
cess. common to multiple specific toxic courses. Each toxic
Figure 6 An overview of TOXic Process InterpretabLe knOwledge sysTem (TOXPILOT) theme is displayed in different colors.
Discussion Using TOXPILOT, researchers can obtain an overall picture of
the mechanism of toxicity in the liver and explore the systemic
There are many biomedical pathway databases, including effects of biological functions. Moreover, from a fragmented
KEGG [27], WikiPathways [28], and Reactome [29]. Since knowledge perspective, our maps facilitate the discovery of new
these databases deal with a large number of pathways, one might knowledge through commonality. Also, our system supports
conclude that they also explain toxic mechanisms. However, both retrospective and forward analyses. In this way,
most of these databases are based on molecular-molecular inter- TOXPILOT enables the generation of knowledge cycles based
actions. Such molecular-centered approaches do not cover cell- on the TXPO (Fig. 6).
or organ-level granularity. The AOP covers key events leading
to adverse effects with varying granularities [30]. However, the Conclusions
AOP focuses primarily on measurable changes. Furthermore, as
an essential point, the AOP is not an ontology, and the terms
described in its pathways lack consistency and in some cases are In the present work, we developed a TXPO to organize toxic
redundant. Ontology can provide richer information flexibly by process knowledge. As an application, we developed the
TOXPILOT as a prototype system for supporting the interpreta-
generalization, specialization, and other relationships in a con-
sistent manner. The TXPO is an ontology and systematizes toxic tion of toxicity mechanisms. We are currently annotating more
processes according to an is-a hierarchy with inheritances from toxic courses and enhancing the level of sophistication of the
terms in the TXPO. In the future, we plan to cover toxic courses
general to specific terms based on a philosophical view that
makes the intrinsic nature explicit. Moreover, by employing in other organs, such as the kidney. We are also planning to re-
systemic functional decomposition, the TXPO covers various use various ontologies, such as the Disease Ontology
(http://disease-ontology.org/), and the Monarch Disease Ontolo-
processes across granularities in a consistent manner. We con-
firmed that we can describe both pathway- and molecular-level gy (Mondo, https://github.com/cmungall/tbd-disease-ontology.)
processes in a unified manner with regard to ER stress. Howev- Bridging domains on toxicity knowledge from basic to clinical
er, we found that the number of molecular processes is so large medicine could help elucidate multiple mechanisms of toxicity.
that it can be difficult to grasp the overall picture of the mecha- In furthering the applications of the TOXPILOT, we are striv-
nism. Therefore, the TXPO deals primarily with process-process ing to enhance its functions. The first version of the TXPO is
interactions with grain sizes from the organelle level. With re- available via the NCBO BioPortal, and the prototype
gard to the molecular level, we describe molecules as partici- TOXPILOT is open at the following site:
pants in toxic course processes. Furthermore, we explain the role https://toxpilot.nibiohn.go.jp. New term requests and reporting
of each molecule in a given specific process. of issues can be made via a GitHub tracker
Understanding toxicity mechanisms is a hard task. Among the (https://github.com/txpo-ontology/TXPO/issues.) We plan to
many issues involved, one aspect is the complexity of various submit TXPO to the OBO foundry for collaboration and
interactions in the toxic course. We demonstrated that our im- knowledge sharing among not only toxicologists but also other
balance model can make the context clearer and distinguish tox- biomedical communities.
ic actions from body defense functions in each granularity, thus
facilitating interpretations of toxic mechanisms. Interestingly, Acknowledgements
we found that sometimes one functioning process plays both a
biological defense role and toxic role. For example, as shown in This research was supported by AMED under grant number
Table 2, during the course of ER stress, apoptosis plays a defen- 19nk0101103h0005. The authors would like to thank Dr. K.
sive role in removing abnormal cells accumulating unfolded Horimoto, Dr. K. Fukui, Prof. Y. Uesawa, and S. Ueda. The
proteins, whereas increasing apoptosis has a toxic effect at the authors also thank Prof. R. Mizoguchi for useful discussions
organ level that can lead to liver failure. Furthermore, our im- related to the ontological approach.
balance model is possible to explain that an imbalance also oc-
curs when defensive functioning becomes excessive. For in-
stance, when the cell proliferation function becomes excessive,
Address for correspondence
liver carcinogenesis can develop at the organ level. We are cur-
Yuki Yamagata y-yamagata@nibiohn.go.jp
rently trying to introduce the imbalance model for other toxic
courses and clarify the relationships between functioning de- Hiroshi Yamada h-yamada@nibiohn.go.jp
Laboratory of Toxicogenomics Informatics, National Institutes
mand and the defense function.
of Biomedical Innovation, Health and Nutrition
The identification of biomarkers for toxicity prediction using 7-6-8 Asagi, Saito, Ibaraki City, Osaka
machine learning techniques is a frequent objective of computa- 567-0085, Japan
tional toxicology research. However, such machine learning ap-
Phone: +81-72-641-9826
proaches often lack accountability. By annotating markers based
on the ontology of TXPO, associating markers with the toxicity
process as a progression of toxicity development, and by visual-
izing them, it is possible to provide accountability for marker References
genes. Therefore, the TXPO and TOXPILOT will contribute to
the enhancement of safety evaluations. Moreover, the general 1. Chen M, Suzuki A, Borlak J, Andrade RJ, Lucena MI.
course map in TOXPILOT provides an indication of causal rela- Drug-induced liver injury: Interactions between drug properties
tionships across various mechanisms of toxicity. Therefore, it and host factors. J Hepatol. 2015;63(2):503-514.
could be used to discover previously unknown relationships and 2. Klassen CD, ed. Casarett & Doull’s Toxicology, the
contribute to the identification of new risks. Basic Science of Poisons. 8th ed.: McGraw-Hill; 2013.
3. Japanese Society of Toxicologic Pathology. New Toxi- 23. Sasajima M, Kitamura Y, Ikeda M, Mizoguchi R.
cologic Histopathology: Nishimura Co; 2017. Japanese. FBRL: a Function and Behavior Representation Language.
4. Educational Committee, the Japanese Society of Toxi- IJCAI. 1995;95:1830-1836.
cology. Toxicology: Asakura Publishing; 2009. Japanese. 24. Horii I. Toxic effect onset and evaluations of medicinal
5. Strayer DS and Rubin E. Rubin's Pathology: Clinico- drugs--horizon for Darwinian toxicological thought. J Toxicol
pathologic Foundations of Medicine. 7th ed.: Wolters Kluwer; Sci. 2010;35(4):425-35.
2015. 25. Mizoguchi R, Kitamura Y, Borgo S. A Unifying Defi-
6. Kaplowitz N, DeLeve LD. Drug-Induced Liver Dis- nition for artifact and Biological Functions. App Ontol.
ease. Academic Press; 2013. 2016;11(2):129-154.
7. Musen MA. Protégé Team. The Protégé Project: A 26. Mizoguchi R, Kozaki K, Kou H, Yamagata Y, Imai T,
Look Back and a Look Forward. AI Matters. 2015 Jun;1(4):4- Waki K, et al. River Flow Model of Diseases. In the Proceedings
12. of the 2nd International Conference on Biomedical Ontology
8. Glimm B, Horrocks I, Motik B, Stoilo G. HermiT: An (ICBO2011). 2011;63-70.
OWL 2 Reasoner. J Autom Reason. 2014;53(3):245-269. 27. Kanehisa M, Furumichi M, Tanabe M, Sato Y, Mor-
9. Arp R, Smith B, Spear AD. Building Ontologies Using ishima K. KEGG: New Perspectives on Genomes, Pathways,
Basic Formal Ontology. Cambridge, MA: The MIT Press; 2015. Diseases and Drugs. Nucleic Acids Res. 2017;45(D1):D353-
10. Whetzel PL, Noy NF, Shah NH, Alexander PR, Nyulas D361.
C, Tudorache T, Musen MA. BioPortal: Enhanced Functionality 28. Slenter DN, Kutmon M, Hanspers K, Riutta A, Wind-
via New Web Services from the National Center for Biomedical sor J, Nunes N, et al. WikiPathways: a Multifaceted Pathway
Ontology to Access and Use Ontologies in Software Applica- Database Bridging Metabolomics to Other Omics
tions. Nucleic Acids Res. 2011 Jul;39(Web Server issue):W541- Research. Nucleic Acids Res. 2018;46(D1):D661-D667.
W545. 29. Fabregat A, Jupe S, Matthews L, Sidiropoulos K, Gil-
11. Mungall CJ, Torniai C, Gkoutos GV, Lewis SE, Haen- lespie M, Garapati P. Garapati, et al. The Reactome Pathway
del MA. Uberon, an Integrative Multi-Species Anatomy Ontolo- Knowledgebase. Nucleic Acids Res. 2018;46(D1):D649-D655.
gy. Genome Biol. 2012;13(1):R5. 30. Ankley GT, Bennett RS, Erickson RJ, Hoff DJ, Hor-
12. Diehl AD, Meehan TF, Bradford YM, Brush MH, nung MW, Johnson RD, et al. Adverse Outcome Pathways: a
Dahdul WM, Dougall DS, et al. The Cell Ontology 2016: En- conceptual Framework to Support Ecotoxicology Research and
hanced Content, Modularization, and Ontology Interoperability. Risk Assessment. Environ Toxicol Chem. 2010;29:730-741.
J Biomed Semantics. 2016;7(1):44.
13. Federhen S. The NCBI Taxonomy Database, Nucleic
Acids Res. 2012;40(Database issue):D136-D143.
14. Hastings J, de Matos P, Dekker A, Ennis M, Harsha B,
Kale N. The ChEBI Reference Database and Ontology for Bio-
logically Relevant Chemistry: Enhancements for 2013. Nucleic
Acids Res. 2013;41(Database issue):D456-D463.
15. Ashburner M, Ball CA, Blake JA, Botstein D, Butler H,
Cherry JM. Gene Ontology: Tool for the Unification of Biology.
The Gene Ontology Consortium. Nat Genet. 2000;25(1):25-29.
16. Gkoutos V, Green C, Mallon M, Hancock., Davidson
D. Using ontologies to describe mouse phenotypes, Genome
Biol., 2005:6 (R8).
17. Yamamoto S, Noriko S, Nakamura H, Fukagawa H,
Fukuda K, Takagi T. INOH: ontology-based highly structured
database of signal transduction pathways, Database, Volume
2011, 2011, bar052.
18. He Y, Liu Y, Zhao B. OGG: a Biological Ontology for
Representing Genes and Genomes in Specific Organisms, in
Proceedings of the 2014 International Conference on Biomedical
Ontologies (ICBO); Houston, TX, USA. 2014:13-20.
19. Apache Jena Fuseki, [cited 2019 Feb 21]. Available
from: http://jena.apache.org/documentation/fuseki2/
20. Bostock M, Ogievetsky V, Heer J. D³: Datadriven
documents. IEEE Trans. Vis Comput. Graph. 2011 17(12):2301-
2309.
21. Smith B, Ashburner M, Rosse C, Bard J, Bug W,
Ceusters W, et al. The OBO Foundry: Coordinated Evolution of
Ontologies to Support Biomedical Data Integration. Nature Bio-
technol. 2007;25(11):1251-1255.
22. Kitamura Y, Mizoguchi R. Ontology-Based Systemati-
zation of Functional Knowledge. J Engineering Design.
2004;15(4):327-351.