Towards a Reference Plant Trait Ontology for Modeling Knowledge
of Plant Traits and Phenotypes
Elizabeth Arnaud
1
, Laurel Cooper
2
, Rosemary Shrestha
3
, Naama Menda
6
, Rex T. Nelson
5
,
Luca Matteis
1
, Milko Skofic
1
, Ruth Bastow
4
, Pankaj Jaiswal
2
, Lukas Mueller
6
and Graham McLaren
7
1
Bioversity International, via dei Tre Denari, 174/a, Maccarese, Rome, Italy
2
Dept. of Botany and Plant Pathology, Oregon State University, Corvallis, OR 97331, U.S.A.
3
Centro Internacional de Mejoramiento de Maiz y Trigo (CIMMYT), Texcoco, Edo. de México, Mexico
4
GARNet, School of Life Sciences, University of Warwick, Wellesbourne, Warwick, CV35 9EF, U.K.
5
USDA-ARS CICGRU, 1012 Crop Informatics and Genetics Laboratory, Ames, IA 50011, U.S.A.
6
Boyce Thompson Institute for Plant Research, Tower Road, Ithaca, NY 14850, U.S.A.
7
Generation Challenge Program, CIMMYT, Texcoco, Edo. de México, Mexico
Keywords: Agriculture, Plant Phenotype, Plant Trait Ontology, Integrated Breeding Knowledge, Community of
Practice.
Abstract: Ontology engineering and knowledge modeling for the plant sciences is expected to contribute to the
understanding of the basis of plant traits that determine phenotypic expression in a given environment.
Several crop- or clade-specific plant trait ontologies have been developed to describe plant traits important
for agriculture in order to address major scientific challenges such as food security. We present three
successful species and/or clade-specific ontologies which address the needs of crop scientists to quickly
access a wide range of trait related data, but their scope limits their interoperability with one another. In this
paper, we present our vision of a species-neutral and overarching Reference Plant Trait Ontology which
would be the basis for linking the disparate knowledge domains and that will support data integration and
data mining across species.
1 INTRODUCTION
Agricultural crop- or clade-specific databases
provide comparative phenotypic and genotypic
information that helps elucidate functional aspects of
plant and agricultural biology. For researchers, it is
necessary to have seamless access to various
distributed and interrelated data sources such as
genetic, trait, genotypic and experimental data to
explore biologically interesting questions.
Agricultural centers have a huge amount of
historical data that reflects a sound scientific
knowledge of crop biology and physiology. Plant
scientists are producing large volumes of data on
genetic mapping, gene expression, and full genome
sequences that can be used to gain better insights
into plant traits and phenotypes.
Traditionally, phenotype information has been
captured in a free text manner, which cannot be
easily indexed and presents an obstacle to data
sharing. One approach to overcome this obstacle is
through the annotation of data using a common
controlled vocabulary or “ontology" (Ashburner et
al., 2000; Smith et al., 2007). An ontology is a way
of representing knowledge in a given domain that
includes a set of terms to describe the classes in that
domain, as well as the relationships among terms.
Each term can be associated with an array of data
such as names, definitions, identification numbers,
and genes involved. Ontologies are fundamental for
unifying diverse terminologies, and are increasingly
used by scientists in many fields and by the online
web search engines. In an ontology, terms are
carefully defined and are related to each other using
logically defined relationships as defined by the
OBO Foundry Relations Ontology (RO; Smith et al.,
2005) and supported by the prevailing knowledge.
Such structured ontology trees allow researchers to
220
Arnaud E., Cooper L., Shrestha R., Menda N., Nelson R., Matteis L., Skofic M., Bastow R., Jaiswal P., Mueller L. and McLaren G..
Towards a Reference Plant Trait Ontology for Modeling Knowledge of Plant Traits and Phenotypes.
DOI: 10.5220/0004138302200225
In Proceedings of the International Conference on Knowledge Engineering and Ontology Development (KEOD-2012), pages 220-225
ISBN: 978-989-8565-30-3
Copyright
c
2012 SCITEPRESS (Science and Technology Publications, Lda.)
use terms consistently in scientific publications or
standardized handbooks on quality/trait evaluations,
and to search for and integrate data linked to these
terms in anatomical, genetic, genomic, and other
types of biological databases.
2 THE SEMANTIC LANDSCAPE
OF GENOTYPE, PHENOTYPE,
AND TRAIT
The concepts of genotype and phenotype are among
the most fundamental in all of genetics,
developmental and evolutionary biology. Plant
breeding, particularly requires the integration of
these concepts to understand how and why
phenotypic expression varies with the environment.
A multidisciplinary approach will help address such
complex questions. Crop modeling can play a
crucial role and requires knowledge integration,
which means that molecular geneticists,
physiologists and crop modelers can share their
respective ‘language’ (Wollenweber, 2005) and
ontology engineering provides a mean to achieve
this.
A genotype of an organism is the inherited
instructions it carries within its genetic code (i.e. the
genome). A genotype can be characterized by
sequencing genes, as well as by genetic mapping to
characterize variations in the DNA sequence. Not all
organisms with the same genotype look or act the
same way because appearance and behavior are
modified by environmental and developmental
conditions. Likewise, not all organisms that look
alike necessarily have the same genotype.
A phenotype (from Greek phainein, 'to show' +
typos, 'type') is the composite of an organism's
observable characteristics such as its morphology,
development, biochemical or physiological
properties, phenology, behavior, and products of
behavior (Wollenweber, 2005). Phenotypes result
from the expression of an organism's genes and
develop over time as the outcome of cumulative
causal interactions between genotype and
environment (Malosetti et al, 2011). The phenotype
comprises the observable characteristics and the
expression of particular traits in a particular
organism or organism part. It is a composite of an
entity (e.g. fruit) and an attribute (e.g. shape) with a
value (e.g. round):
Entity + Attribute = Trait
(1)
Entity + (Attribute + Value) =
Phenotype (observed)
(2)
Example:
fruit + (shape + round) = fruit shape round
-> round fruit
3 ENGINEERING COMMON
SEMANTIC FRAMEWORK - A
REFERENCE PLANT TRAIT
ONTOLOGY
Complex free text descriptions used for phenotypes
are almost impossible to index and retrieve in a
useful way unless ontological concepts are used for
the metadata and text tagging. The semantic problem
is that, depending on the plant species, the same trait
can be given different names. For instance, the trait
term seed color is referred to as kernel color in
maize, grain or caryopsis color in rice and pod color
in beans but these are all based on the same
phenotypic descriptor- fruit color. Data integration
and/or mining of plant trait data require the
identification of equivalent concepts used by the
various agricultural research communities.
Having phenotype data scattered in various
online databases using their own vocabularies for
annotation prevents the integration and comparison
of phenotypic and genetic data between species and
even across taxa.
The solution to this problem is the development
of a true Reference Plant Trait Ontology (Ref-TO;
Figure 1). The basis of the proposed Ref-TO is the
existing Plant Trait Ontology (TO) (Figure 2) which
will integrate and link many crop- and clade-specific
trait ontologies. Initially the focus will be on
integrating the three crop or clade-specific
ontologies described below, but the long-term goal
is to describe traits of all plant species.
The development and expansion of such a
universal ontology will rely on the long-term
involvement of the various plant research
communities for maintaining the species-specific
terms and applying the ontological terms to their
data annotations.
TowardsaReferencePlantTraitOntologyforModelingKnowledgeofPlantTraitsandPhenotypes
221
Figure 1: Vision of a Reference Plant Trait Ontology (Ref-TO) to link the crop- and clade-specific trait ontologies.
4 SPECIFIC ONTOLOGIES
PROVIDE PLANT PHENOTYPE
AND TRAIT INFORMATION
The agricultural research community requires crop
trait information, which may contribute to the
comparative analysis of genomes and to the
selection of promising plant material by crop
breeders.
4.1 Soybean Trait Ontology (SOY)
A controlled vocabulary (ontology) describing
soybean traits is under active development (SOY),
as part of SoyBase (http://soybase.org; Grant et al,
2010), the USDA-ARS soybean genetic and
genomic database, a professionally curated
biological database for soybean genetic and
molecular data. This controlled vocabulary uses
terms familiar to the soybean community to facilitate
its use. Genetic markers, QTL and soybean gene
data are linked to the SOY controlled vocabulary but
also cross-referenced to the Plant Trait Ontology
(TO) for extension to other crop species. Queries can
be initiated at the SoyBase portal using SOY, TO
and Plant Ontology (PO) identifiers to access
soybean data regarding soybean traits and/or
anatomical structures.
In anticipation of further development of the TO,
web services are also being developed to allow
programmatic access to soybean data using soybean
(SOY), Plant Trait (TO), Plant Ontology (PO) or
Gene Ontology (GO) identifiers. Semantic web
queries of SoyBase data are also available using
SSWAP (Gessler et al. 2009) services (Nelson et al.
2010).
4.2 Solanaceae Phenotype Ontology
(SPO)
As part of the community-driven SOL Genomics
Network (SGN;
http://solgenomics.net; Bombarely
et al., 2011, Menda 2008), an ontology for
Solanaceae phenotypes (SPO) has been developed to
describe traits and phenotypes scored by plant
breeders in the field. The SGN is a clade-oriented
comparative genomic database, focusing on the
Euasterid clade, including the Solanaceae family,
which has many important crop and model plants
such as tomato (Solanum lycopersicum), potato
(Solanum tuberosum), eggplant (Solanum
melongena), and pepper (Capsicum annuum).
Since many Solanaceae phenotype ontology
terms are pre-composed, these are also mapped to
one or more Plant Ontology (PO) terms, and the
Phenotype Quality Ontology (PATO) terms (e.g. the
SPO term ‘late fruit ripening’ is mapped to the
PATO term ‘delayed’ and PO growth stage term
‘ripening’ and plant anatomy term ‘fruit’).
4.3 The Crop Ontology (CO)
The Crop Ontology (CO) (http://
www.cropontology.org/) of the Generation
Challenge Programme (GCP) aims to provide a
semantic framework to the computational
architecture of the knowledge-based system called
KEOD2012-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
222
the Integrated Breeding Platform (https://
www.integratedbreeding.net/).
The CO is designed to provide a structured,
controlled vocabulary for the phenotype of
important crops for food and agriculture and is
collectively developed by various Crop
Communities, associated with the centers of the
Consultative Group on International Agricultural
Research.
The aim is to foster consistency in annotation
and to aggregate datasets containing huge amount of
historic phenotypic data on a large range of crops
not adequately represented in the PO and the TO
(Shrestha et al., 2011). Crops currently included are:
banana, cassava, common bean, cowpea, groundnut,
maize, potato, rice, sorghum, soybean, wheat.
Barley, pigeon pea and yam will be added in 2013.
The CO describes agronomic, morphological,
physiological, quality, and abiotic and biotic stress
related traits of several crops using a number of
common relationship types. However, relations were
created such as ‘method_ of’, ‘scale_of’, and
’derived_from’ to meaningfully describe the traits
and their relations to methods and scales.
The CO contributes to the expansion of the Plant
Ontology (PO) and to the Plant Trait Ontology (TO),
through submission of new terms. Web links
between CO terms cross-referenced with major
agronomic information sources provide online
access to data annotated with similar ontological
terms.
The online Crop Ontology is a public resource
that acts as an open-source server for names of traits.
5 TOWARDS THE REFERENCE
TRAIT ONTOLOGY
5.1 The Trait Ontology (TO) and the
Plant Ontology (PO)
International collaborative efforts already exist to
develop multi-species ontologies for example the
Plant Ontology (PO) and Trait Ontology (TO). The
TO (Figure 2) describes phenotypic traits in plants.
In its current form, the TO is organized around eight
main classes, allowing it to encompass a broad range
of plant traits and be species-neutral. The TO is
being actively developed in close cooperation with
the Plant Ontology, which describes the
morphological and anatomical structures of all
plants, as well as the stages of development of the
plant structures. (Avraham et al, 2008, Jaiswal,
2005, Walls et al, 2012).
TO terms are “precomposed” (Entity-Quality
(EQ) form) using terms from the PO and the Gene
Ontology (GO), along with other ontologies such as
the Phenotype Quality Ontology (PATO), the
ontology of Chemical Entities of Biological Interest
(ChEBI), the Plant Environmental Conditions
Ontology (EO) and the Plant Disease Ontology
(PDO; currently under development), as well as
others.
The PO itself has been extensively revised over
the past two years, with the focus on expanding the
scope of the ontology to span all green plants. We
can apply the lessons learned from the PO
development to developing the TO as a reference
trait ontology for all plant species. Further
development of the TO is necessary to develop the
across-species terms that will be useful to
semantically link the crop- and clade- specific
ontologies to one another.
5.2 Common Platforms for Data
Integration through Web Services
All the crop- and clade-specific ontologies, as well
as the PO and TO, are being developed using a
common platform, the OBO-Edit software (Day-
Richter et al, 2007) developed and promoted by the
Gene Ontology (GO; The Gene Ontology
Consortium, 2010). This facilitates cross-linking.
All these ontologies also use a number of common
relationship types. The most common are ‘is_a’ and
part_of’ relations assigned by OBO-foundry (Smith
et al, 2005).
The ontologies presented in this paper are
available on the BioPortal site of the National Center
for Biomedical Ontology (NCBO) (http://
bioportal.bioontology.org) for public access, as well
as on their respective sites.
Via various processors or extractors, Resource
Description Framework (RDF; http://
www.w3.org/RDF/) can capture and convey the
metadata or information in unstructured (e.g. text),
semi-structured (e.g. HTML documents) or
structured sources (e.g. standard databases). This
makes RDF a perfect solution for representing data
that exists in various databases. RDF structures
enable synonyms or aliases to be easily mapped to
the same types or concepts. This kind of semantic
matching is a key capability of the semantic Web.
Currently, both the Plant Ontology and Crop
Ontology are available in RDF.
TowardsaReferencePlantTraitOntologyforModelingKnowledgeofPlantTraitsandPhenotypes
223
Figure 2: A model of the existing Plant Trait Ontology (TO) showing the species-neutral approach and interaction with the
Gene Ontology (GO), the Phenotype Quality Ontology (PATO), the ontology of Chemical Entities of Biological Interest
(ChEBI), the Plant Environmental Conditions Ontology (EO) and the Plant Disease Ontology (PDO).
The final objective of a programmatic use of a
trait ontology is to support the integration of data
sets for given traits, retrieval through web services
and the discovery of any piece of information that is
annotated with analogous trait concepts. Currently,
the existence of many distinct ontologies results in a
discontinuous semantic framework. Each ontology is
presently taking further steps to use web services to
synchronize trait names and OBO files. For
example, the GCP crop databases and field books for
breeders are synchronized for data annotations with
CO through the API.
Developers who wish to use the Plant Ontology
in mobile or desktop applications can now access
terms, synonyms, definitions, and comments using
PO web services. Built with PHP (http://
www.php.net/) and modelling aspects of RESTful
software architecture (Fielding, 2000), these services
provide PO data encoded in JavaScript Object
Notation (JSON) format, a widely-used standard for
providing data over the internet. The PO plans to
continue to develop these web services and
envisions the Reference Plant Trait Ontology being
offered in a similar way in the future.
6 CONCLUSIONS
A Reference Plant Trait Ontology is necessary to
unify the clade- and crop- specific ontologies and
provide the semantic framework for querying,
reasoning and data mining across the various species
databases. Therefore, our objective is to develop the
Reference Plant Trait Ontology by improving and
expanding the existing Plant Trait Ontology. Our
vision for the future development of the Ref-TO is
one of an international consortium of the clade- and
crop-specific trait ontologies, and would also include
representatives from the model plant database
groups (such as GARNet, NASC and TAIR for
Arabidopsis) and representatives from the Plant
Ontology and Gene Ontology.
Cross-referencing species-specific terms will
unite the ontologies into a network and, by linking
plant phenotypes and traits to information, images
and documentation across species and even taxa, the
community is building a knowledge base with a
broad reach, which will be useful to elucidate
functional aspects of plant and agricultural biology.
REFERENCES
Avraham, S., Tung, C.-W., Ilic, K., Jaiswal, P., Kellogg,
E. A., McCouch, S., Pujar, A., et al. (2008). The Plant
Ontology Database: a community resource for plant
structure and developmental stages controlled
vocabulary and annotations. Nucleic Acids Research,
36(suppl_1), D449–454.
Bombarely, A., Menda, N., Tecle, I.Y., Buels, R.M.,
Strickler, S., Fischer-York, T., Pujar, A., et al. (2011).
The Sol Genomics Network (solgenomics.net):
growing tomatoes using Perl. Nucleic Acids Research,
39(suppl 1).
Day-Richter, J., Harris, M.A., Haendel, M., The Gene
Ontology OBO Edit Working Group, and Lewis, S.
KEOD2012-InternationalConferenceonKnowledgeEngineeringandOntologyDevelopment
224
(2007). Obo-Edit – An ontology editor for biologists.
Bioinformatics 23:2198–2200.
Fielding, Roy T. (2000). Architectural styles and the
design of network-based software architectures (Book,
2000) [WorldCat.org]. University of California, Irvine.
Gessler, D. D. G., Schiltz, G.S., May, G. D., Avraham, S.,
Town, C. D., Grant, D., and Nelson R. T. (2009).
SSWAP: A Simple Semantic Web Architecture and
Protocol for semantic web services. BMC
Bioinformatics 10:309
Grant, D., Nelson, R. T., Cannon, S.B. and Shoemaker,
R.C. (2010). SoyBase, the USDA-ARS soybean
genetics and genomics database. Nucl. Acids Res. 38
(Suppl.1):D843-D846.
Jaiswal, P., Avraham, S., Ilic, K., Kellogg, E.A.,
McCouch, S., Pujar, A., Reiser, L., et al. (2005). Plant
Ontology (PO): a controlled vocabulary of plant
structures and growth stages. Comparative and
Functional Genomics, 6(7-8), 388–397.
Jaiswal P, 2011. Gramene Database: A Hub for
Comparative Plant Genomics Plant Reverse Genetics
Methods in Molecular Biology, Volume 678, 247-275,
Jaiswal, P., Ware, D., Ni, J., Chang, K., Zhao, W.,
Schmidt, S., Pan, X., et al. (2002). Gramene:
development and integration of trait and gene
ontologies for rice. Comparative and Functional
Genomics, 3(2), 132–136.
Malesotti M., Ribaut JM, Eeuwijk van E (2011). The
statistical analysis of multienvironment data:
modelling genotype-by-environement interaction and
its genetic basis. In Drought phenotyping in crops:
from theory to practice, Part I Plant phenotyping
methodology. 123-145
Menda, N., Buels, R. M., Tecle, I., & Mueller, L. A.
(2008). A Community-Based Annotation Framework
for Linking Solanaceae Genomes with Phenomes.
Plant Physiology, 147(4), 1788–1799.
Nelson, R., Avraham, S., Shoemaker, R., May, G., Ware,
D., & Gessler, D. (2010). Applications and methods
utilizing the Simple Semantic Web Architecture and
Protocol (SSWAP) for bioinformatics resource
discovery and disparate data and service integration.
BioData Mining, 3(1), 3.
Shrestha, R., Arnaud, E., Mauleon, R., Senger, M.,
Davenport, G. F., Hancock, D., Morrison, N., et al.
(2010). Multifunctional crop trait ontology for
breeders’ data: field book, annotation, data discovery
and semantic enrichment of the literature. AoB Plants,
2010.
Shrestha, R., Davenport, G. F., Bruskiewich, R., and
Arnaud, E. (2011). Development of crop ontology for
sharing crop phenotypic information. In Drought
phenotyping in crops: from theory to practice, eds. P.
Monneveux and J.M. Ribaut (Generation Challenge
Programme (GCP), c/o CIMMYT, Mexico) 167-176.
Smith, B., Ashburner, M., Rosse, C., Bard, J., Bug, W.,
Ceusters, W., Goldberg, L. J., et al. (2007). The OBO
Foundry: coordinated evolution of ontologies to
support biomedical data integration. Nature
Biotechnology, 25(11), 1251–1255.
Smith B, Ceusters W, Klagges B, Kohler J, Kumar A,
Lomax J, Mungall CJ, Neuhaus F, Rector A, Rosse C.,
2005 Relations in biomedical ontologies. Genome
Biology;6(5):R46.
Walls, R.L., Athreya, B., Cooper, L., Elser, J., Gandolfo,
M.A., Jaiswal, P., Mungall, C.J., et al. (2012).
Ontologies as integrative tools for plant science.
American Journal of Botany, In press.
Wollenweber B, Porter JR, Lubberstedt T (2005) Need for
multidisciplinary research towards a second green
revolution. Curr Opin Plant Biol 8: 337-341
TowardsaReferencePlantTraitOntologyforModelingKnowledgeofPlantTraitsandPhenotypes
225