ABI Development: An ontology of evidence types to support biological data management

点击翻译按钮获取中文摘要

Researchers generate biological data from many diverse methods that range from laboratory experiments to computer-based analyses. These data serve as the evidence that researchers use to make inferences and draw scientific conclusions. The process of biocuration seeks to capture these conclusions and the evidence that led to them in a standardized way so that the information is readily accessible to the entire scientific community. The most efficient way to accomplish this is to use an ontology to describe the evidence types. An ontology is a controlled vocabulary of terms where each term is carefully defined and linked to other terms by precise relationships. The Evidence Ontology (EO) is a community standard for describing types of research evidence used to support scientific conclusions in biological research. The EO is used by some of the world?s most prominent protein databases and genomic resources to capture evidence information. The goal of this project is to improve EO and promote its use by a larger community of researchers. The EO will be promoted through outreach, training, and education efforts, including workshops and internships. Broader impacts will include outreach efforts to Baltimore City Public Schools students focusing on teaching the importance of structuring information in a controlled way. Summer interns will engage in EO development and bioinformatics activities. A vast number of scientists researching a wide range of biological topics will benefit from the continued development and expanded use of the EO.The ability to describe both evidence and assertion method (i.e. whether a human or a machine makes a statement) in a consistent and computable fashion is essential for multiple reasons. Capture of methodology is central to the scientific method and can impact evaluation of results, associating structured evidence with stored data allows for selective data queries and retrieval from even the largest databases, and structured evidence systems make automated quality control possible, which is essential for large-scale data management. Nearly 30 biological resources including protein databases, model organism databases, phenotype resources, and gene expression databases currently are using the Evidence Ontology (EO) to capture evidence information, support structured data queries, group related data, or establish quality control mechanisms. EO will be developed further to address structural issues, clarify the main axis, add logical constraints, and map EO to related resources. New evidence types will be continually added to the ontology based on the needs of the research community. A web resource will be created that includes improved visualization tools for evidence and data associated with EO terms, complete user documentation, and downloadable content. EO will also develop quality assessment methodologies to enable researchers to better evaluate evidence. Outreach, training, and education will be conducted to grow the EO user base and educate researchers and students about the value and means of capturing evidence. EO developers will present at scientific conferences, publish papers, host interns, and conduct workshops and science outreach activities. By improving EO and increasing user awareness, researchers will be better able to make the most of evidence and associated data. For more information, please visit: http://evidenceontology.org.