Chapter Four - The Neurobehavior Ontology: An Ontology for Annotation and Integration of Behavior and Behavioral Phenotypes

https://doi.org/10.1016/B978-0-12-388408-4.00004-6Get rights and content

Abstract

In recent years, considerable advances have been made toward our understanding of the genetic architecture of behavior and the physical, mental, and environmental influences that underpin behavioral processes. The provision of a method for recording behavior-related phenomena is necessary to enable integrative and comparative analyses of data and knowledge about behavior. The neurobehavior ontology facilitates the systematic representation of behavior and behavioral phenotypes, thereby improving the unification and integration behavioral data in neuroscience research.

Introduction

The study of the behavior of organisms forms a major biological discipline encompassed via the investigation of physical, mental, and environmental influences that underpin behavioral-related processes. Geneticists have been studying behavior since 1800s when Francis Galton started investigating heredity and human behavior systematically (Rose & Rose, 2011). We now know that one of the most important factors for behavioral variation within and across organisms lies in genetic diversity (Hamer, 2002, Mackay, 2008). Behavioral geneticists attempt to unravel this behavioral variation by investigating the underlying mechanisms that govern it in an effort to elucidate our understanding of the pathogenesis of neuropsychiatric disorders (Congdon, Poldrack, & Freimer, 2010).

The great successes and advances both in genomics and in our abilities to quantify and analyze genomic information have transformed genetics over the past decade. Behavioral geneticists take advantage of these in order to gain an in-depth understanding of the genetic architecture of behavior. They seek to understand what genes affect behavior, how they interact with other genes, what is the molecular basis of their allelic variation, and how this variation behaves with respect to the environment (Holden, 2001). One of the tools that they employ to achieve these goals is the use of animal models that provide a platform where complex behaviors can be studied and quantified with substantial progress over the past in recent years, especially in respect with research related to the mouse and the fruit fly Drosophila (Mackay, 2008, Wehner et al., 2001).

Animal models have been proven useful for unveiling the genetic basis of many behavior-related diseases including various neurodegenerative disorders such as Parkinson's, Huntington's, spinocerebellar ataxia, and Alzheimer's disease, as well as for providing the medium for novel drug discovery. Furthermore, animal models for diseases whose indicators are formed by behavioral observations rather than definitive neuropathological markers are being developed. For example, there are various mouse models of loss of Fragile X mental retardation 1 (Fmr1) or methyl-CpG-binding protein-2 (Mecp2) or ubiquitin protein ligase 3A (Ube3A) function that underlie syndromes associated with autistic-like behavior (Moy & Nadler, 2007).

There are now large international projects, consortia, and individual labs around the world that study and record the effect of genetic variations in various species and at various levels of granularity. Behavioral screens are part of the assays performed and include the study of a variety of behavioral phenotypes such as reproductive behavior, learning and memory, feeding behavior, sleep, and circadian rhythm (Brown et al., 2005, Levin and Cerutti, 2009, Sokolowski, 2001, Spuhler, 2009, Tecott and Nestler, 2004). The resulting data provide us with a wealth of information that can be exploited to investigate and reveal the molecular basis of behavior and behavioral disorders. However, while other domains of biology have made significant progress in systematically structuring and analyzing their data, we do not currently have a standardized way to characterize behavioral processes and phenotypes (Congdon et al., 2010). The provision of a method for recording behavior-related phenomena is necessary to enable integrative and comparative analyses of data and knowledge about behavior (Gkoutos, Green, Mallon, Hancock, & Davidson, 2004b).

In other areas of biomedical science, similar demands have lead to the generation of various resources that allow for the systematic characterization, organization, and recording of knowledge and data (Schofield, Sundberg, Hoehndorf, & Gkoutos, 2011a). In particular, the advent of the gene ontology (GO) (Ashburner et al., 2000) has provided a critical landmark in the use of ontologies to harmonize the description of domains of knowledge and facilitated the development of several other ontologies for various different domains. Ontologies are structured, standardized terminologies in which some aspect of the meaning of terms has been rendered computable. For example, the GO does not only include textual definitions of its terms but also machine readable, computable relations (such as is-a, part-of, and regulates) that enable the automated traversal of the ontology and analysis of the underlying data. Perhaps more importantly, the standardization of the terminology and the inclusion of computable definitions paved the way for interoperability between biomedical databases and have lead to the possibility for large-scale integration of biomedical data (Bada et al., 2004, Chen et al., 2012, Hoehndorf et al., 2012, Hoehndorf et al., 2011, Hoehndorf et al., 2011).

Here, we present our efforts toward creating a framework that allows the systematic representation of behavior processes and related phenotype manifestations offering the tantalizing possibility of unifying behavioral data across species integrating and translating our knowledge so as to provide new grounds for targeting behavior-related diseases.

Section snippets

Neurobehavior ontology

Understanding what constitutes behavior will depend on its formal definition and the systematic representation of the processes involved in behavioral mechanisms. According to Tinbergen (1963), behavior biology is primarily concerned with four major questions: causation (mechanism), development (ontogeny), function (adaptation), and evolution (phylogeny) (Adcock, 2001). These four questions can be collapsed into two categories—the proximate (“how”) category that includes causation and

Human behavior phenotypes

Dissecting the genetic basis of behavior variation in humans is an important factor toward our understanding of human disease. The potential to identify the molecular underpinnings of human behavior and its characteristics depends on our ability to make meaningful genotype–phenotype correlations. Behavioral manifestations recorded in the clinic are not only an invaluable diagnostic tool but also provide insights to human pathophysiology and pathobiology. For example, the distinct behavioral

Discussion

The NBO is one of the first comprehensive ontologies designed for the integration of behavioral observations in animal organisms and humans. NBO's prime application is to provide the vocabulary that is required to integrate behavior observations within and across species. It is currently being applied by several model organism communities as well as for the description of human behavior-related disease phenotypes, and the use of a common, shared vocabulary for data annotation will lead to the

Ontology

The initial version of the ontology was developed using a combination of OBO-edit (Richter, Harris, Haendel, & Lewis, 2007) and emacs. Subsequently, we transformed the ontology into the OWL format and it is currently maintained using Protege4 (Noy et al., 2001). In addition to simple relationships connecting classes, NBO contains a wide range of additional logical axioms, which are intended primarily assist with automated maintenance, quality control, and classification of the ontology.

NBO and phenotype ontologies

Acknowledgments

This work was supported by the National Institutes of Health (Grant number R01 HG004838-02) and the European Commission's 7th Framework Programme, RICORDO project (Grant number 248502).

References (67)

  • A. Abbott

    Mouse project to find each gene's role

    Nature

    (2010)
  • J. Adcock

    Animal behavior: An evolutionary approach

    (2001)
  • J. Amberger et al.

    A new face and new challenges for online Mendelian inheritance in man (OMIM)

    Human Mutation

    (2011)
  • M. Ashburner et al.

    Gene ontology: Tool for the unification of biology

    Nature Genetics

    (2000)
  • J. Bolhuis et al.

    Animal behaviour

    (2009)
  • Y. Bradford et al.

    ZFIN: Enhancements and updates to the zebrafish model organism database

    Nucleic Acids Research

    (2011)
  • S.D. Brown et al.

    EMPReSS: Standardized phenotype screens for functional annotation of the mouse genome

    Nature Genetics

    (2005)
  • C.J. Bult et al.

    The mouse genome database (MGD): Integrating biology with the genome

    Nucleic Acids Research

    (2004)
  • S.B. Cassidy et al.

    Behavioral phenotypes in genetic syndromes: Genetic clues to human behavior

    Advances in Pediatrics

    (2002)
  • M.A. Cenci et al.

    Animal models of neurological deficits: How relevant is the rat?

    Nature Reviews. Neuroscience

    (2002)
  • C.-K. Chen et al.

    Mousefinder: Candidate disease genes from mouse phenotype data

    Human Mutation

    (2012)
  • A. Codita et al.

    Of mice and men: More neurobiology in dementia

    Current Opinion in Psychiatry

    (2006)
  • V.P. Dole et al.

    Toward an analogue of alcoholism in mice: Analysis of nongenetic variance in consumption of alcohol

    Proceedings of the National Academy of Sciences of the United States of America

    (1988)
  • R. Drysdale

    Phenotypic data in FlyBase

    Briefings in Bioinformatics

    (2001)
  • R. Drysdale et al.

    FlyBase: A database for the drosophila research community

    Methods in Molecular Biology (Clifton, N.J.)

    (2008)
  • D.A. Finn et al.

    Genetic animal models of anxiety

    Neurogenetics

    (2003)
  • K.L. Gilby

    A new rat model for vulnerability to epilepsy and autism spectrum disorders

    Epilepsia

    (2008)
  • G.V. Gkoutos et al.

    Using ontologies to describe mouse phenotypes

    Genome Biology

    (2004)
  • G.V. Gkoutos et al.

    Building mouse phenotype ontologies

    Pacific Symposium on Biocomputing

    (2004)
  • G.V. Gkoutos et al.

    Building mouse phenotype ontologies

  • G.V. Gkoutos et al.

    Using ontologies to describe mouse phenotypes

    Genome Biology

    (2005)
  • G.V. Gkoutos et al.

    Entity/quality-based logical definitions for the human skeletal phenome using PATO

  • J. Grier et al.

    Biology of animal behavior

    (1992)
  • Cited by (31)

    • Reproducibility and replicability of rodent phenotyping in preclinical studies

      2018, Neuroscience and Biobehavioral Reviews
      Citation Excerpt :

      Should an “immobility” label be used rather than labels such as “anxiety”, “fear,” “learning” and “behavioral despair? Efforts such as the Neurobehavioral Ontology (Gkoutos et al., 2012), the Vertebrate Trait Ontology (Park et al., 2013) and the Mammalian Phenotype Ontology (Smith and Eppig, 2012) have each taken different approaches to this issue based in large part on their unique applications. In databases storing multiple phenotypic measures from multiple studies and laboratories, the implications of these approaches for replicability and reproducibility may for the first time be methodically investigated.

    • A knowledge-based system for generating interaction networks from ecological data

      2017, Data and Knowledge Engineering
      Citation Excerpt :

      Ontologies have successfully been used to standardise metadata terminology for the discovery, integration (by semantic mediation), and re-use of biodiversity and ecological data [28]. Limited modelling of behavioural concepts has been undertaken e.g. in the context of human neurobiology [29,30]. In biodiversity and ecosystem informatics (BDEI) aspects of behavioural ecology have been modelled in male jumping-spider courtship behaviour, sea-turtle nesting behaviour [31] and the behaviour of social insects [32].

    • A knowledge-based system for discovering ecological interactions in biodiversity data-stores of heterogeneous specimen-records: A case-study of flower-visiting ecology

      2014, Ecological Informatics
      Citation Excerpt :

      The Neuro Behavior Ontology contains classes of two fundamental types, namely BehavioralProcess and BehavioralPhenotype, the sub-classes of which constitute a species-independent behavior vocabulary that is interoperable with the Gene Ontology and with species-specific phenotype ontologies such as those of the human, mouse, fly and worm. One of the main objectives of developing the Neuro Behavior Ontology is to discover the genetic basis of disease (Gkoutos et al., 2012). In BDEI the use of ontology modeling in the study of behavior, including behavioral ecology, has been addressed in an ontology of male jumping-spider courtship behavior and an ontology of sea turtle nesting behavior (Midford, 2004).

    View all citing articles on Scopus
    View full text