Semantic Search for Biomedical Texts using Predicate-Argument

Structure

Mohammed Alliheedi

and Robert E. Mercer

Department of Computer Science, Al Baha University, Prince Mohammad Bin Saud, Al Bahah 65527, Saudi Arabia

Department of Computer Science, The University of Western Ontario, 1151 Richmond St., London, Ontario, Canada

Keywords:

Biomedical Ontology, Semantic Search, Semantic Roles, Knowledge Representation, Formal Ontology.

Abstract:

In this position paper we argue that using semantic roles in addition to using biologically-oriented ontologies

and databases (or knowledge bases) will further enhance the generation of RDF triples that can be collected

from biomedical text. RDF triples have been used to enhance semantic search beyond the simple use of

linguistically oriented additions such as synonyms. We wish to focus on drug-virus interactions.

1 INTRODUCTION

In this position paper we argue that using seman-

tic roles in addition to using linguistically-oriented

and biologically-oriented ontologies and databases

(or knowledge bases) will further enhance the gen-

eration of RDF triples that can be collected from

biomedical text. RDF triples can then be used to

enhance semantic search beyond the simple use of

linguistically oriented additions such as synonyms

(Kohlschein et al., 2018). Linguistically-oriented on-

tologies such as WordNet (Miller, 1995) can provide

synonyms, hyponyms, hypernyms, and meronyms.

VerbNet (Kipper-Schuler, 2005; Kipper et al., 2008)

and BioFrameNet (Dolbey et al., 2006), an im-

portant biologically-oriented extension to FrameNet

(FrameNet, 2020), provides means to connect verbs

with semantic roles. BioFrameNet also provides con-

nections to biological ontologies, such as, GO (Ash-

burner et al., 2000) and Entrez Gene (Entrez Gene,

2020). Recently, interest to extend VerbNet (Kipper-

Schuler, 2005; Kipper et al., 2008) into the biolog-

ical domain has reemerged with BioVerbNet (Chiu

et al., 2019), which follows on the earlier work of

Lippincott et al. (Lippincott et al., 2013), and a sepa-

rate nascent investigation of biochemistry experimen-

tal method procedure verbs (Alliheedi et al., 2019b;

Alliheedi and Mercer, 2019; Alliheedi et al., 2019a;

Alliheedi et al., 2019c).

We wish to focus on drug-virus interactions and

are particularly interested in searching the COVID-19

Open Research Dataset (Wang et al., 2020) for drug-

virus interactions and how the drugs interact with the

viruses. While these just mentioned linguistic ontolo-

gies provide important information about the verbs in

the form of semantic roles that the verbs take as lin-

guistic arguments, what is missing are the relations

among biological entities. These relationships can be

captured by machine accessible RDF triples, subject-

verb-object triples that use the standard nomenclature

provided by various terminology ontologies. Triples

in the biology domain are collected by tools such as

those found in Bio2RDF.org (Bio2RDF, 2020). This

type of machine-accessible knowledge is strongly ar-

gued for in the FAIR guidelines (Wilkinson et al.,

2016). Our immediate objective is to use the seman-

tic role information found in the linguistic ontologies

to generate the RDF triples found in scientiﬁc pub-

lications and the text sections found in many of the

biological knowledge bases. Our long term goal is

to use these RDF triples in a semantic search engine

targeted on the COVID-19 Open Research Dataset.

2 ONTOLOGIES AND

KNOWLEDGE SOURCES

In this section, we give an overview about exist-

ing biologically oriented ontologies in the litera-

ture. Table 1 shows some of well known ontolo-

gies that have been developed over the last decades.

These ontologies include the Gene Ontology (GO)

(Ashburner et al., 2000), the Foundational Model of

Anatomy (FMA) (Rosse and Mejino Jr, 2003), the

Relation Ontology (RO) (Smith et al., 2005), the

Alliheedi, M. and Mercer, R.

Semantic Search for Biomedical Texts using Predicate-Argument Structure.

DOI: 10.5220/0010150702990306

In Proceedings of the 12th International Joint Conference on Knowledge Discovery, Knowledge Engineering and Knowledge Management (IC3K 2020) - Volume 2: KEOD, pages 299-306

ISBN: 978-989-758-474-9

299

Table 1: List of developed ontologies in the literature.

Ontology name Domain Developed by

The Gene Ontology (GO)

describes knowledge of the gene

domain.

(Ashburner et al., 2000;

Gene Ontology Consortium, 2011)

The Foundational Model of

Anatomy Ontology (FMA)

is an ontology about human anatomy (Rosse and Mejino Jr, 2003)

The Relation Ontology (RO)

consists of relations intended to be

used across various ontologies in OBO

Foundry

(Smith et al., 2005)

The Ontology of Scientiﬁc

Experiments (EXPO)

deﬁnes the general knowledge about

scientiﬁc experimental aspects (e.g.,

methodology and design)

(Soldatova and King, 2006)

The Ontology for Chemical

Entities of Biological

Interest (ChEBI)

a database of molecular entities

focusing on small molecules

(Degtyarenko et al., 2007)

The Molecular Methods

Database (MolMeth)

a resource consists of scientiﬁc

protocol ontologies

(Klingstr

om et al., 2013)

The Ontology of Scientiﬁc

Experiments (EXACT)

describes scientiﬁc protocols and

experiments

(Soldatova et al., 2013)

Semanticscience Integrated

Ontology

describes scientiﬁc protocols and

experiments

(Dumontier et al., 2014)

The Ontology for

Biomedical Investigations

(OBI)

a resource for annotating biomedical

investigations

(Bandrowski et al., 2016)

ontology of Scientiﬁc Experiments (EXPO), the on-

tology for Biomedical Investigations (OBI) (Solda-

tova and King, 2006), the ontology for Chemical En-

tities of Biological Interest (ChEBI), the ontology

of scientiﬁc experiments (EXACT) (Soldatova et al.,

2013), and the Molecular Methods Database (Mol-

Meth) (Klingstr

om et al., 2013). These ontologies are

discussed brieﬂy in this section. Most of these on-

tologies describe a set of concepts and categories in

the biological domain that shows their properties and

the relations between them. The goal of these ontolo-

gies is to provide deﬁnitive controlled terminologies

that describe entities in the biomedical genre. At-

tention towards designing and building ontologies has

become increasingly central in the biomedical domain

(Rosse and Mejino Jr, 2003). The main aspect of GO

is to provide information that describes gene prod-

ucts using a precisely deﬁned vocabulary (Ashburner

et al., 2000). GO was built using various resources

such as those in (FlyBase Consortium, 2003; Blake

et al., 2000; Ringwald et al., 2000; Ball et al., 2000).

Similarly to GO, ChEBI (Degtyarenko et al., 2007)

was created using data from several resources such

as IntEnz (Fleischmann et al., 2004), KEGG COM-

POUND (Kanehisa et al., 2006), and the Chemical

Ontology. While Go and ChEBI focus on Gene and

molecular entities, OBI

(Bandrowski et al., 2016), on

http://purl.obolibrary.org/obo/obi

the other hand, focuses on the annotations of biomed-

ical investigations and provides standard tools to rep-

resent study design, protocols and instrumentation

used, the data generated, and the types of analysis

performed on the data. Several ontologies (Courtot

et al., 2008; Brinkman et al., 2010; Zheng et al., 2013;

Soldatova et al., 2013; Dumontier et al., 2014) were

developed and based on the OBI ontology.

EXPO provides detailed descriptions of various

aspects of scientiﬁc experiments and their relation-

ships (Soldatova and King, 2006) .

Descriptions of experimental processes are pro-

vided by OBI, and three real-world applications are

discussed in (Brinkman et al., 2010). Some of the

relations in these applications (e.g., inputs, outputs)

add to the ontologies that describe biological entities.

The beta cell genomics application ontology (BCGO)

(Zheng et al., 2013) also uses OBI, but it tends to be a

more descriptive ontology than some of the others that

use OBI, but some of the relations in RO, the relation

ontology (Smith et al., 2005), that are used (e.g., pro-

duces, translate to) do have an ordering sense.

EXACT (Soldatova et al., 2013) and the Se-

manticscience Integrated Ontology (Dumontier et al.,

2014) describe scientiﬁc protocols and experiments.

MolMeth is a database which contains scientiﬁc

protocol ontologies that conform to a set of laboratory

protocol standards (Klingstr

om et al., 2013).

Other ontologies describe general concepts that

KEOD 2020 - 12th International Conference on Knowledge Engineering and Ontology Development

300

Table 2: List of Semantic Roles developed in the work of (Alliheedi and Mercer, 2019).

Semantic role Deﬁnition

Agent Generally a human or an animate subject.

Patient Participants that have undergone a process.

Predicate A word that initiates the frame.

Theme Participants in a location or undergoing a change of location.

Goal Identiﬁes a thing toward which an action is directed or a place to which something

moves.

Factitive A referent that results from the action or state identiﬁed by a verb.

Location The physical place where the experiments took place.

Instrument an object or force that comes in contact with an object and causes some change in them

are useful to a biochemistry procedure-oriented on-

tology include: Ontologies consist of process such as

(Lenat et al., 1985) and (Schlenoff et al., 2000), on-

tology for units of measure (Rijgersberg et al., 2013),

classiﬁcation of scenarios and plans (CLASP) (De-

vanbu and Litman, 1996), and materials ontology

(Ashino, 2010). Foundational theories such as pro-

cess calculus and regular grammar are essential for

the formalization of procedure-oriented ontologies.

We consulted three knowledge sources in our dis-

cussion in Section 3. MedChemExpress (MCE)

, a

company, offers a wide range of high quality research

chemicals and biochemicals. Its website provided

synonym information for some biological elements.

GeneCards

is a database of information on all known

and predicted human genes. UniProt

is a database of

protein sequence and functional information.

3 SEMANTIC ROLES

Semantic roles are deﬁned as “the underlying re-

lationships that participants have with a verb in a

clause”

. Minsky deﬁned frame as “a data-structure

representing a stereotyped situation” (Minsky, 1974),

with frames having a header, slots and slot ﬁllers. Fill-

more (Fillmore, 1976) introduced the notion of frame

semantics as a theory of meaning. A semantic frame

is deﬁned by Fillmore as “any coherent individuatable

perception, memory, experience, action or object”. In

other words, these are coherently world events or ex-

periences. In this work, we are interested in develop-

ing frame semantics at the verb level so that our head-

ers are verbs and our slots are semantic roles, ﬁlled

by the words which represent these roles. For exam-

ple, to understand the word “buy”, one would access

the knowledge contained in the commercial transac-

https://www.medchemexpress.com

https://www.genecards.org

https://www.uniprot.org

https://glossary.sil.org/term/semantic-role

tion frame which includes words such as the person

who buys the goods (buyer), the goods that are being

sold (goods), the person who sells the goods (seller),

and the currency that the buyer and seller agree on

(money). Motivated by Fillmore’s theory of frame

semantics, FrameNet (Baker et al., 1998) was devel-

oped to create an online lexical resource for English.

This framework includes more than 170,000 manu-

ally annotated sentences and 10,000 words. The com-

putational linguistic community has been attracted to

the concept of frame semantics and developed com-

putational resources using this concept, such as Verb-

Net (Kipper-Schuler, 2005), an on-line verb lexicon

for English and PropBank (Palmer et al., 2005), an

annotated corpus with basic semantic propositions.

Since we are focusing on drug-virus interaction

verbs with the associated semantic roles, we are par-

ticularly interested in analyzing the COVID-19 Open

Research Dataset (Wang et al., 2020). Verbs evoke se-

mantic roles in writing. Semantic roles provide salient

pieces of information about experimental steps. Se-

mantic roles are crucial aspects of identifying relevant

information in the biomedical texts. This relevant in-

formation is essential to generate the RDF triples (i.e.,

subject, predicate, and object) which have been used

in the literature (Hu et al., 2017). We can use the

work by (Alliheedi and Mercer, 2019) to label each

sentence with the appropriate semantic roles (see Ta-

ble 2). We have used an annotation scheme for iden-

tifying the structured representation of knowledge in

a set of sentences of biochemistry articles (Alliheedi

and Mercer, 2019). The list of semantic roles include:

Theme, Patient, Agent, Location, Goal. These seman-

tic roles identify the arguments of both verbs (e.g.,

administered) and nominalised verbs (e.g., inhibitor

is a nominalized form of the verb inhibit). Further-

more, the use of ontology that describes drug-virus

interactions is salient because it provides the relations

among various molecules.

Semantic Search for Biomedical Texts using Predicate-Argument Structure

301

Figure 1: Steps of the example to ﬁnd information about GS-331007, a metabolite of sofosbuvir.

4 EXAMPLE

We use as an example, a paper that discusses the anti-

viral drug sofosbuvir (Kirby et al., 2015). The title

and abstract of this paper states that sofosbuvir is an

inhibitor of the virus. This produces a very high-level

RDF triple. We would like to have triples that de-

scribe this inhibition in more detail. What is the bio-

chemistry that causes this inhibition? Reading further,

we encounter some information about how this inhibi-

tion takes place. We ﬁrst ﬁnd mention of two metabo-

lites of sofosbuvir, GS-461203 and GS-331007. But

the abstract is not that informative regarding why

these metabolites are mentioned. In the full text of

the paper we ﬁnd “GS-461203, the pharmacologically

active nucleoside analog triphosphate metabolite of

sofosbuvir, is incorporated into HCV RNA by NS5B

polymerase, where it acts as a chain terminator.” So,

this provides us with the information that we need for

the GS-461203 metabolite. We remain interested in

following up with GS-331007. We will need to look

at other information sources. Figure 1 provides a vi-

sual presentation of the search that we now describe.

We cannot ﬁnd mention of GS-331007 in some of

the knowledge bases that we search. However, we

ﬁnd in MedChemExpress (MCE) that PSI-6206 and

GS-331007 are synonyms

. Now searching for PSI-

6206, it is found in GeneCards under the POLR2A

https://www.medchemexpress.com/PSI-6206.html

Gene

. We note information about PSI-6206 and its

importance as an inhibitor of the Hepatitis virus. But

no other information is forthcoming. So, we then

turned to UniProt

. While looking in UniProt under

POLR2A gene under Homo Sapien

we ﬁnd “(Micro-

bial infection) Acts as an RNA-dependent RNA poly-

merase when associated with small delta antigen of

Hepatitis delta virus, acting both as a replicate and

transcriptase for the viral RNA circular genome.” and

a link to a publication (Chang et al., 2008). In this

publication we ﬁnd important information about how

the hepatitis delta virus uses host RNA polymerase

which is the protein encoded by the POLR2A gene.

“Previous studies have indicated that the replication

of the RNA genome of hepatitis delta virus (HDV)

involves redirection of RNA polymerase II (Pol II), a

host enzyme that normally uses DNA as a template.

. . . Taken together, we have demonstrated that with

a low concentration of amanitin that only inhibited

Pol II transcription and did not affect host Pol I or

Pol III transcription there was a signiﬁcant inhibition

for HDV genomic and antigenomic RNAs. Thus, we

believe that Pol II is required for the transcription of

both genomic and antigenomic HDV RNAs.” (Chang

et al., 2008).

https://www.genecards.org/Search/Keyword?

queryString=PSI-6206

https://www.uniprot.org

https://www.uniprot.org/uniprot/P24928

KEOD 2020 - 12th International Conference on Knowledge Engineering and Ontology Development

302

We now have a number of sources that have pro-

vided factual information and textual information.

The factual information (e.g., sofosbuvir is a drug)

can be readily made into an RDF triple. Knowledge

about drugs (drugs ending in ir are anti-viral drugs)

can be used to further reﬁne the triple. The textual

information can be analyzed using the semantic roles

that we have discussed in Section 3. The semantic

roles can be used to form RDF triples (agent is the

subject, predicate is the verb, and theme and instru-

ment are the objects). As an example, we can la-

bel “GS-461203, the pharmacologically active nucle-

oside analog triphosphate metabolite of sofosbuvir, is

incorporated into HCV RNA by NS5B polymerase...”

using the semantic roles in Table 2 as follows:

GS-461203/Patient; is incorporated/Predicate; into

HCV RNA/Theme; by NS5B polymerase/Agent.

This example produces a number of RDF triples:

Triples

– sofosbuvir is-a drug

– sofosbuvir is-a anti-viral drug

– hepatitis delta is-a virus

– sofosbuvir inhibits Hepatitis

– GS-461203 is-a metabolite of sofosbuvir

– hepatitis C virus is-a-synonym-of HCV

– HCV NS5B polymerase incorporates GS-461203

into HCV RNA

– GS-461203 acts-as chain terminator of HCV RNA

synthesis

– GS-331007 is-a metabolite of sofosbuvir

– hepatitis delta virus is-a-synonym-of HDV

– POLR2A gene encodes RNA polymerase

– RNA polymerase is-essential-for cell function

– RNA polymerase acts-as an RNA-dependent RNA

polymerase

– RNA polymerase associates-with small delta anti-

gen of Hepatitis delta virus

– RNA polymerase acts-as replicate for the viral RNA

circular genome

– RNA polymerase acts-as transcriptase for the viral

RNA circular genome

– RNA polymerase II is-a RNA polymerase

– RNA polymerase II is-required-for transcription of

genomic HDV RNA

– RNA polymerase II is-required-for transcription of

antigenomic HDV RNA

5 SEMANTIC SEARCH

Standard information retrieval has relied on the vec-

tor space model (Salton et al., 1975). The vec-

tor space model for representing documents in high-

dimensional vector space has been validated by

decades of research and development. This model

represents documents and queries as vectors (based

in various ways, such as the almost universally used

TF-IDF (Salton et al., 1975), on the words occurring

in the document and the query) and then using some

measure (typically, the cosine difference between the

query and each document vector) to retrieve relevant

documents. Although very popular these types of rep-

resentation of document semantics based solely on

ﬁrst order document-term statistics, such as TF-IDF,

are limited in their expressiveness and search recall.

Semantic search (or semantic information re-

trieval) started in the early days of Artiﬁcial Intelli-

gence (Raphael, 1964). There have been two main

directions to improve the vector space model: 1)

improve the vector representation of the query and

the document in order to improve the relationship

between the query and the relevant document vec-

tors, and 2) provide meta-knowledge, typically in the

form of synonyms, hyponyms, and hypernyms, but

sometimes with the more distantly related meronyms,

homonyms, or semantic ﬁelds, in general. Some of

the methods that have been developed are directly ap-

plicable to semantic search.

The typical techniques used are to include syn-

onyms, hyponyms, and hypernyms from the WordNet

ontology (Fern

andez et al., 2011). Although we will

investigate this knowledge source, we will also use

the biomedical ontologies (Bandrowski et al., 2016;

Courtot et al., 2008; Brinkman et al., 2010; Zheng

et al., 2013; Soldatova et al., 2013; Dumontier et al.,

2014; Soldatova and King, 2006; Zheng et al., 2013)

to build our own ontologies (Liang et al., 2017; Seit-

ner et al., 2016) because the specialized biomedical

domains have specialized meanings for many com-

mon words and many technical terminologies that re-

quire specialized ontologies. In addition to these lin-

guistically motivated semantic search extensions, we

will also use RDF triples that have been derived from

biomedical texts and information sources using our

suggested application of semantic roles.

As an example of how a query with additional se-

mantic information is used in a semantic search:

Query: How does sofosbuvir help infected-patients

with hepatitis C virus?

A search request is processed using the following pro-

cedure: The user search query is analyzed with lin-

guistic tools and is matched against the word ontolo-

gies and the RDF data source, retrieving synonyms

and RDF descriptions that semantically match each

entity in the query with semantic information encoded

in the RDF triples. A full text search query of various

databases, based on the data returned by the previ-

Semantic Search for Biomedical Texts using Predicate-Argument Structure

303

ous step, is generated. All resulting information is

returned to the user.

For this example query, the search would return

papers and other information that talks about how so-

fosbuvir works to inhibit hepatitis viruses, such as the

drug acting to prevent hepatitis C RNA synthesis.

6 CONCLUSION AND FUTURE

WORK

The long-term goal that has motivated this position

paper is to provide a semantic search system to query

the COVID-19 Open Research Dataset (Wang et al.,

2020). We have argued that RDF triples can assist this

semantic search. Moving toward this goal, our imme-

diate objective is to use the previously presented se-

mantic roles to assist the automatic generation of the

RDF triples found in scientiﬁc publications and the

text sections found in many of the biological knowl-

edge bases. This step together with information de-

rived from ontologies and knowledge bases are nec-

essary for building a semantic search system that is

capable of extracting the relevant information from

biomedical texts that goes beyond simple keyword

searches.

Our proposed idea is the ﬁrst step towards devel-

oping an automated semantic search. We aim to reﬁne

the methods discussed in this paper and to develop a

semantic search system. We intend to focus on the

drug-virus interactions (e.g., how different antiviral

drugs interact with COVID-19). A few drug interac-

tion ontologies exist

. A knowledge base with virus-

drug interactions exists

. Having the RDF triples de-

rived from text and this knowledge base would allow

us to move toward a drug-virus ontology.

REFERENCES

Alliheedi, M. and Mercer, R. E. (2019). Semantic roles:

Towards rhetorical moves in writing about experimen-

tal procedures. In Proceedings of the 32nd Canadian

Conference on Artiﬁcial Intelligence, pages 518–524.

Alliheedi, M., Mercer, R. E., and Cohen, R. (2019a). Anno-

tation of rhetorical moves in biochemistry articles. In

Proceedings of the 6th Workshop on Argument Min-

ing, pages 113–123.

Alliheedi, M., Mercer, R. E., and Haas-Neill, S. (2019b).

Ontological knowledge for rhetorical move analysis.

Computaci

on y Sistemas, 23(3):633–647.

https://bioportal.bioontology.org

https://go.drugbank.com

Alliheedi, M., Wang, Y., and Mercer, R. E. (2019c). Bio-

chemistry procedure-oriented ontology: A case study.

In Proceedings of the 11th International Conference

on Knowledge Engineering and Ontology Develop-

ment (KEOD 2019) – Volume 2 of Proceedings of the

11th International Conference on Knowledge Discov-

ery, Knowledge Engineering and Knowledge Manage-

ment (IC3K 2019), pages 164–173.

Ashburner, M., Ball, C. A., Blake, J. A., Botstein, D.,

Butler, H., Cherry, J. M., Davis, A. P., Dolinski, K.,

Dwight, S. S., Eppig, J. T., et al. (2000). Gene ontol-

ogy: tool for the uniﬁcation of biology. Nature Genet-

ics, 25(1):25.

Ashino, T. (2010). Materials ontology: An infrastructure

for exchanging materials information and knowledge.

Data Science Journal, 9:54–61.

Baker, C. F., Fillmore, C. J., and Lowe, J. B. (1998). The

Berkeley FrameNnet project. In Proceedings of the

17th International Conference on Computational Lin-

guistics - Volume 1, pages 86–90.

Ball, C. A., Dolinski, K., Dwight, S. S., Harris, M. A., Issel-

Tarver, L., Kasarskis, A., Scafe, C. R., Sherlock, G.,

Binkley, G., Jin, H., et al. (2000). Integrating func-

tional genomic information into the saccharomyces

genome database. Nucleic Acids Research, 28(1):77–

80.

Bandrowski, A., Brinkman, R., Brochhausen, M., Brush,

M. H., Bill Bug and, M. C. C., Clancy, K., Cour-

tot, M., Derom, D., Dumontier, M., Fan, L., Fos-

tel, J., Fragoso, G., Gibson, F., Gonzalez-Beltran, A.,

Haendel, M. A., He, Y., Heiskanen, M., Hernandez-

Boussard, T., Jensen, M., Lin, Y., Lister, A. L., Lord,

P., Malone, J., Manduchi, E., Monnie McGee and,

N. M., Overton, J. A., Parkinson, H., Peters, B.,

Rocca-Serra, P., Ruttenberg, A., Sansone, S.-A.,

Scheuermann, R. H., Schober, D., Smith, B., Solda-

tova, L. N., Christian J. Stoeckert, J., Taylor, C. F.,

Torniai, C., Turner, J. A., Vita, R., Whetzel, P. L., and

Zheng, J. (2016). The ontology for biomedical inves-

tigations. PLoS ONE, 11(4):e0154556.

Bio2RDF (2020). https://bio2rdf.org.

Blake, J. A., Eppig, J. T., Richardson, J. E., Davisson, M. T.,

Group, M. G. D., et al. (2000). The mouse genome

database (mgd): expanding genetic and genomic re-

sources for the laboratory mouse. Nucleic Acids Re-

search, 28(1):108–111.

Brinkman, R. R., Courtot, M., Derom, D., Fostel, J. M.,

He, Y., Lord, P., Malone, J., Parkinson, H., Peters,

B., Rocca-Serra, P., Ruttenberg, A., Sansone, S.-A.,

Soldatova, L. N., Jr., C. J. S., Turner, J. A., Zheng, J.,

and the OBI consortium (2010). Modeling biomedical

experimental processes with OBI. Journal of Biomed-

ical Semantics, 1 (Suppl 1):S7.

Chang, J., Nie, X., Chang, H. E., Han, Z., and Taylor, J.

(2008). Transcription of Hepatitis delta virus RNA by

RNA polymerase II. Journal of Virology, 82(3):1118–

1127.

Chiu, B., Majewska, O., Pyysalo, S., Wey, L., Stenius,

U., Korhonen, A., and Palmer, M. (2019). A neu-

ral classiﬁcation method for supporting the creation

KEOD 2020 - 12th International Conference on Knowledge Engineering and Ontology Development

304

of BioVerbNet. Journal of Biomedical Semantics,

10(2):12.

Courtot, M., Bug, W., Gibson, F., Lister, A. L., Malone,

J., Schober, D., Brinkman, R. R., and Ruttenberg, A.

(2008). The OWL of biomedical investigations. In

Proceedings of the Fifth OWLED Workshop on OWL:

Experiences and Directions, page 12pp.

Degtyarenko, K., De Matos, P., Ennis, M., Hastings, J.,

Zbinden, M., McNaught, A., Alc

antara, R., Dar-

sow, M., Guedj, M., and Ashburner, M. (2007).

ChEBI: a database and ontology for chemical enti-

ties of biological interest. Nucleic Acids Research,

36(suppl 1):D344–D350.

Devanbu, P. T. and Litman, D. J. (1996). Taxonomic plan

reasoning. Artiﬁcial Intelligence, 84(1-2):1–35.

Dolbey, A., Ellsworth, M., and Scheffczyk, J. (2006).

BioFrameNet: A domain-speciﬁc FrameNet exten-

sion with links to biomedical ontologies. In Second In-

ternational Workshop on Formal Biomedical Knowl-

edge Representation (KR-MED 2006), CEUR Work-

shop Proceedings Vol. 222, pages 87–94.

Dumontier, M., Baker, C. J., Baran, J., Callahan, A., Che-

pelev, L., Cruz-Toledo, J., Rio, N. R. D., Duck, G.,

Furlong, L. I., Keath, N., Klassen, D., McCusker,

J. P., Queralt-Rosinach, N., Samwald, M., Villanueva-

Rosales, N., Wilkinson, M. D., and Hoehndorf, R.

(2014). The semanticscience integrated ontology (sio)

for biomedical research and knowledge discovery.

Journal of Biomedical Semantics, 5(1):14.

Entrez Gene (2020). https://www.ncbi.nlm.nih.gov.

Fern

andez, M., Cantador, I., L

opez, V., Vallet, D., Castells,

P., and Motta, E. (2011). Semantically enhanced

Information Retrieval: An ontology-based approach.

Journal of Web Semantics, 9(4):434–452.

Fillmore, C. J. (1976). Frame semantics and the nature of

language. Annals of the New York Academy of Sci-

ences, 280(1):20–32.

Fleischmann, A., Darsow, M., Degtyarenko, K., Fleis-

chmann, W., Boyce, S., Axelsen, K. B., Bairoch,

A., Schomburg, D., Tipton, K. F., and Apweiler,

R. (2004). Intenz, the integrated relational enzyme

database. Nucleic Acids Research, 32(suppl 1):D434–

D437.

FlyBase Consortium (2003). The ﬂybase database of the

drosophila genome projects and community literature.

Nucleic Acids Research, 31(1):172–175.

FrameNet (2020). https://framenet.icsi.berkeley.edu.

Gene Ontology Consortium (2011). The gene ontology:

enhancements for 2011. Nucleic Acids Research,

40(D1):D559–D564.

Hu, W., Qiu, H., Huang, J., and Dumontier, M. (2017).

BioSearch: a semantic search engine for Bio2RDF.

Database, 2017.

Kanehisa, M., Goto, S., Hattori, M., Aoki-Kinoshita, K. F.,

Itoh, M., Kawashima, S., Katayama, T., Araki, M.,

and Hirakawa, M. (2006). From genomics to chemical

genomics: new developments in kegg. Nucleic Acids

Research, 34(suppl 1):D354–D357.

Kipper, K., Korhonen, A., Ryant, N., and Palmer, M.

(2008). A large-scale classiﬁcation of English verbs.

Language Resources and Evaluation, 42(1):21–40.

Kipper-Schuler, K. (2005). VerbNet: A broad-coverage,

comprehensive verb lexicon. PhD thesis, University

of Pennsylvania.

Kirby, B. J., Symonds, W. T., Kearney, B. P., and Mathias,

A. A. (2015). Pharmacokinetic, pharmacodynamic,

and drug-interaction proﬁle of the hepatitis C virus

NS5B polymerase inhibitor sofosbuvir. Clinical Phar-

macokinetics, 54(7):677–690.

Klingstr

om, T., Soldatova, L., Stevens, R., Roos, T. E.,

Swertz, M. A., M

uller, K. M., Kala

s, M., Lambrix,

P., Taussig, M. J., Litton, J.-E., Landegren, U., and

Bongcam-Rudlof, E. (2013). Workshop on labortory

protocol standards for the molecular methods data-

base. New Biotechnology, 30(2):109–113.

Kohlschein, C., Klischies, D., Paulus, A., Burgdorf, A.,

Meisen, T., and Kipp, M. (2018). An extensible se-

mantic search engine for biomedical publications. In

2018 IEEE 20th International Conference on e-Health

Networking, Applications and Services (Healthcom),

pages 1–6.

Lenat, D. B., Prakash, M., and Shepherd, M. (1985). Cyc:

Using common sense knowledge to overcome brittle-

ness and knowledge acquisition bottlenecks. AI mag-

azine, 6(4):65–65.

Liang, J., Zhang, Y., Xiao, Y., Wang, H., Wang, W., and

Zhu, P. (2017). On the transitivity of hypernym-

hyponym relations in data-driven lexical taxonomies.

In Proceedings of the Thirty-First AAAI Conference

on Artiﬁcial Intelligence, AAAI’17, page 1185–1191.

AAAI Press.

Lippincott, T., Rimell, L., Verspoor, K., and Korhonen,

A. (2013). Approaches to verb subcategorization

for biomedicine. Journal of Biomedical Informatics,

46(2):212–227.

Miller, G. A. (1995). Wordnet: A lexical database for en-

glish. Communications of the ACM, 38(11):39–41.

Minsky, M. (1974). A framework for representing knowl-

edge. Artiﬁcial Intelligence Memo No. 306, Mas-

sachusetts Institute of Technology A.I. Laboratory.

Palmer, M., Gildea, D., and Kingsbury, P. (2005). The

Proposition Bank: An annotated corpus of semantic

roles. Computational Linguistics, 31(1):71–106.

Raphael, B. (1964). SIR: A computer program for seman-

tic information retrieval. PhD thesis, Massachusetts

Institute of Technology, Cambridge, USA.

Rijgersberg, H., Van Assem, M., and Top, J. (2013). Ontol-

ogy of units of measure and related concepts. Seman-

tic Web, 4(1):3–13.

Ringwald, M., Eppig, J. T., Kadin, J. A., and Richardson,

J. E. (2000). Gxd: a gene expression database for the

laboratory mouse: current status and recent enhance-

ments. Nucleic Acids Research, 28(1):115–119.

Rosse, C. and Mejino Jr, J. L. (2003). A reference on-

tology for biomedical informatics: the foundational

model of anatomy. Journal of Biomedical Informat-

ics, 36(6):478–500.

Salton, G., Wong, A., and Yang, C. S. (1975). A vector

Semantic Search for Biomedical Texts using Predicate-Argument Structure

305

space model for automatic indexing. Commun. ACM,

18(11):613–620.

Schlenoff, C., Schlenoff, C., Tissot, F., Valois, J., and Lee,

J. (2000). The Process Speciﬁcation Language (PSL)

overview and version 1.0 speciﬁcation. Internal Re-

port (NISTIR) - 6459, NIST Interagency.

Seitner, J., Bizer, C., Eckert, K., Faralli, S., Meusel, R.,

Paulheim, H., and Ponzetto, S. P. (2016). A large

DataBase of hypernymy relations extracted from the

Web. In Proceedings of the Tenth International

Conference on Language Resources and Evaluation

(LREC’16), pages 360–367.

Smith, B., Ceusters, W., Klagges, B., K

ohler, J., Kumar, A.,

Lomax, J., Mungall, C., Neuhaus, F., Rector, A. L., ,

and Rosse, C. (2005). Relations in biomedical ontolo-

gies. Genome Biology, 6(5):R46.

Soldatova, L., King, R., Basu, P., Haddi, E., and Saunders,

N. (2013). The representation of biomedical proto-

cols. EMBnet.journal, 19(B).

Soldatova, L. N. and King, R. D. (2006). An ontology of

scientiﬁc experiments. Journal of the Royal Society

Interface, 3(11).

Wang, L. L., Lo, K., Chandrasekhar, Y., Reas, R., Yang, J.,

Eide, D., Funk, K., Kinney, R. M., Liu, Z., Merrill,

W., Mooney, P., Murdick, D. A., Rishi, D., Sheehan,

J., Shen, Z., Stilson, B., Wade, A. D., Wang, K., Wil-

helm, C., Xie, B., Raymond, D. M., Weld, D. S., Et-

zioni, O., and Kohlmeier, S. (2020). CORD-19: The

covid-19 open research dataset. ArXiv.

Wilkinson, M. D., Dumontier, M., Aalbersberg, I. J., Apple-

ton, G., Axton, M., Baak, A., Blomberg, N., Boiten,

J.-W., da Silva Santos, L. B., Bourne, P. E., Bouw-

man, J., Brookes, A. J., Clark, T., Crosas, M., Dillo,

I., Dumon, O., Edmunds, S., Evelo, C. T., Finkers, R.,

Gonzalez-Beltran, A., Gray, A. J., Groth, P., Goble,

C., Grethe, J. S., Heringa, J., ’t Hoen, P. A., Hooft,

R., Kuhn, T., Kok, R., Kok, J., Lusher, S. J., Martone,

M. E., Mons, A., Packer, A. L., Persson, B., Rocca-

Serra, Philippeand Roos, M., van Schaik, R., Sansone,

S.-A., Schultes, E., Sengstag, T., Slater, T., Strawn,

G., Swertz, M. A., Thompson, M., van der Lei, J.,

van Mulligen, E., Velterop, J., Waagmeester, A., Wit-

tenburg, P., Wolstencroft, K., Zhao, J., and Mons, B.

(2016). The FAIR Guiding Principles for scientiﬁc

data management and stewardship. Scientiﬁc Data,

3(1):9.

Zheng, J., Manduchi, E., and Stoeckert Jr, C. J. (2013). De-

velopment of an application ontology for beta cell ge-

nomics based on the ontology for biomedical investi-

gations. In 4th International Conference on Biomedi-

cal Ontology, pages 62–67.

KEOD 2020 - 12th International Conference on Knowledge Engineering and Ontology Development

306