The flora phenotype ontology (FLOPO): tool for integrating morphological traits and phenotypes of vascular plants

Abstract

BackgroundThe systematic analysis of a large number of comparable plant trait data can support investigations into phylogenetics and ecological adaptation, with broad applications in evolutionary biology, agriculture, conservation, and the functioning of ecosystems. Floras, i.e., books collecting the information on all known plant species found within a region, are a potentially rich source of such plant trait data. Floras describe plant traits with a focus on morphology and other traits relevant for species identification in addition to other characteristics of plant species, such as ecological affinities, distribution, economic value, health applications, traditional uses, and so on. However, a key limitation in systematically analyzing information in Floras is the lack of a standardized vocabulary for the described traits as well as the difficulties in extracting structured information from free text.ResultsWe have developed the Flora Phenotype Ontology (FLOPO), an ontology for describing traits of plant species found in Floras. We used the Plant Ontology (PO) and the Phenotype And Trait Ontology (PATO) to extract entity-quality relationships from digitized taxon descriptions in Floras, and used a formal ontological approach based on phenotype description patterns and automated reasoning to generate the FLOPO. The resulting ontology consists of 25,407 classes and is based on the PO and PATO. The classified ontology closely follows the structure of Plant Ontology in that the primary axis of classification is the observed plant anatomical structure, and more specific traits are then classified based on parthood and subclass relations between anatomical structures as well as subclass relations between phenotypic qualities.ConclusionsThe FLOPO is primarily intended as a framework based on which plant traits can be integrated computationally across all species and higher taxa of flowering plants. Importantly, it is not intended to replace established vocabularies or ontologies, but rather serve as an overarching framework based on which different application- and domain-specific ontologies, thesauri and vocabularies of phenotypes observed in flowering plants can be integrated.Electronic supplementary materialThe online version of this article (doi:10.1186/s13326-016-0107-8) contains supplementary material, which is available to authorized users.

Extraction of phenotypic traits from taxonomic descriptions for the tree of life using natural language processing

“…For example, currently some substructures like “base” cannot be related to their parent structure, such as “tree” or “leaf.” Therefore, the user may not be able to discern if the information the ETC pipeline extracted under the “size of the base” describes the “base of the plant” or “base of the leaf.” By incorporating ontologies in the NLP pipeline, the software can create a bridge between “base” and its parent structure (i.e., plant or leaf). In addition to extracting phenotypic characters that can be used in the exploration of the plant tree of life, the terms extracted and traits generated using this pipeline will help expand other infrastructures that seek to make terms comparable, inferable, and searchable (e.g., Plant Ontology ‘PO’: Jaiswal et al., ; Planteome Project [http://www.planteome.org]: Cooper and Jaiswal, ; Flora Phenotype Ontology: Hoendorf et al., ) so that phenotypic data sets can be incorporated in analyses in a variety of biological fields.…”

“…Taxonomic descriptions often describe a broader range of character traits, including both qualitative and quantitative traits that provide a summary of the variation observed within a taxon (e.g., length of leaf: 6–10 cm; shape of leaf: ovate to obovate). Consequently, recent research has focused on developing the infrastructure, including software, glossaries, and ontologies, to automate the large‐scale extraction of phenotypic data from taxonomic descriptions (Jaiswal et al., ; Cui, ; Burleigh et al., ; Hamman et al., ; Garnier et al., ; Hoendorf et al., ; Endara et al., ).We describe a natural language processing (NLP) pipeline that leverages this new infrastructure to build character‐by‐taxon phenotypic trait matrices that are usable for evolutionary inference from formal taxonomic descriptions written in English. The NLP pipeline uses a non‐supervised learning strategy that analyzes the full length of the body of a description.…”

The anatomy of phenotype ontologies: principles, properties and applications

“…For example, character matrices will represent phenotypes of individual organisms or species, without comparing them a priori to other entities. Similarly, in a biodiversity context, such as floras or monographs focusing on the organisms within a region, or characterizing a family of related organisms, phenotype descriptions are those of individual organisms or species, without including a comparison [ 60 ].…”

“…These general rules can be used to automatically generate a PATO-based backbone taxonomic structure of a phenotype ontology. For example, such an automated approach has been applied in generating prototypes of the Flora Phenotype Ontology [ 60 ] or the Cellular Phenotype Ontology [ 120 ]. Additionally, for manually generated phenotype ontologies, formal PATO-based ontology definitions can be created for some or all classes and the background knowledge in anatomy and physiology ontologies can be used to improve the ontology quality.…”

Ontology-based support for taxonomic functions

“…• FLOPO, the Flora Phenotype Ontology [87], [88], [88]. FLOPO mostly focuses on supporting the analysis of organismic traits specificallt on the plant community, by establishing a standardized vocabulary describing traits of plant species found in Flora; • HPO, the Human Phenotype Ontology [89] aims to establish a standardized, controlled vocabulary that allows human phenotype information to be described in an unambiguous fashion in medical publications and databases.…”