posted on 2018-12-05, 12:57authored byŞenay Kafkas, Marwa Abdelhakim, Maxat Kulmanov, Marwa Abdellatif, Paul Shofield, Robert Hoehndorf
Accurate and rapid identification of pathogens that are causative for a set of patient phenotypes is crucial for the treatment of serious infectious disease and to determine specific and effective drugs.
Both traditional and metagenome-based microbial diagnostic methods have limitations in identifying causative pathogens or distinguishing causative from non-causative micro-organisms. Phenotype-based methods have the potential to provide an independent source of information in selecting causative pathogens; comparing phenotypes associated with
pathogens to the phenotypes observed in a patient can contribute to selecting and filtering causative pathogens.
We have developed PathoPhenoDB, a database containing
pathogen-to-pheno\-type associations for supporting research of mechanisms in infectious diseases. PathoPhenoDB relies on manual curation of pathogen-disease relations, and on ontology-based text mining to associate phenotypes with infectious diseases.
PathoPhenoDB currently contains phenotypes associated with 692 pathogens and 508 infectious diseases. Using Semantic Web technologies, PathoPhenoDB is enriched with the known drug resistance genes from the pathogens available in the Antibiotics Resistance Ontology and drugs used in treatment of infectious diseases. PathoPhenoDB presents a unique source of information for infectious disease diagnosis support by providing the background knowledge based on searches on known pathogens, diseases and phenotypes. PathoPhenoDB is accessible at http://patho.phenomebrowser.net, and all data is freely available through a public SPARQL endpoint.
Funding
King Abdullah University Science and Technology Office of Sponsored Research (OSR) under Award No.~URF/1/3454-01-01 and FCC/1/1976-08-01.