Background Plant pathogens are one of the main causes of crop losses worldwide, and for that reason are subject of intensive study in many research centers. Although the association between plants and bacteria in nature can be benefitial (as in the case of plant-growth promoting rhizobacteria, also known as PGPR), it is well established that several bacterial genera can attack and cause disease in many plant species, some of them being economically important crops. One of the best characterized bacterial genus is Pseudomonas. This genus contains some members that have a non-pathogenic lifestyle and others that exhibit plant pathogenic activity, like the extensively studied bacterium Pseudomonas syringae. But apart from these genus, there is evidence for a large set of other bacteria genera that present plant-pathogenic activity (http://www.ncppb.com), affecting many different plant species. The idea behind all these is that the amount of biological data related with this research field is very extensive and sometimes scattered and difficult to manage and integrate.
Life Science Semantic Web This technology is the result of applying the semantic technologies such as RDF or OWL to solve the issue of knowledge management in the field of life sciences. It can be described as a newtwork of data, ontologies and their associated services that can be accessed to infer new knowledge in an automatic way. As of today, certain life science projects make use of this technology and therefore new life science-related tools have been developed: the Gene Ontology (http://www.geneontology.org/), Bio2RDF and many others. Related to plant biotechnology, these technologies have been applied less extensively; some examples are the Plant Ontology (http://www.plantontology.org/) developed by the Plant Ontology Consortium; the Plant Trait Ontology (http://www.gramene.org/plant_ontology/) or the Plant Disease Ontology. In the domain of plant-pathogenic bacteria, these technologies have not been implemented so far.
Goal In order to integrate the previous tools in the field of plant pathogenic organisms, we have created Plant Pathogen Interactions Ontology, an ontology that establishes the first step towards the axiomisation of plant-pathogen interactions; this platform offers the scaffold into which important data related with this domain can be embedded.
PPIO is being constructed from a plant-pathogen point of view, in order to complement other knowledge bases such as Plant Ontology or the Plant Trait Ontology. The design of this novel ontology sets the ground for further modelling, being its main structure expressive enough for modeling the main aspects that a plant pathologist is looking forward to express. The development of PPIO is being automated as much as possible, and many ontological concepts are programatically produced using a bioinformatics-oriented workflow platform called Galaxy. In addition, we are now pursuing a knowledge capture project aimed at collecting relevant, manually-verified data, that will ultimately populate the PPIO knowledge base. The main goal is to use PPIO to build an ontology that constitutes a generic Plant - Pathogen Interactions knowledge model, that encourages consistent annotation and supports both query and inference. We hope to achieve the objective of making PPIO an essential bioinformatics tool for the plant-pathogen community. Recently, we contacted with the PHI consortium (http://www.phi-base.org/) and obtained permission to use their data to populate PPIO. We have extracted the plant-pathogen related datasets and we are now developing the tools and the different modelling approaches, in order to integrate this knowledge into the PPIO semantic structure. This will convert PPIO into a more data-enriched ontology, providing new information like virulence-related genes, etc.