Assessing the quality of annotations in asthma gene expression experiments

BMC Bioinformatics

Proceedings

Ronilda Lacson¹  Lucila Ohno-Machado²  Michael Mbagwu³  Hisham Yousif⁴ 

[1] Decision Systems Group, Brigham & Women’s Hospital, Harvard Medical School, Boston, MA, USA;Division of Biomedical Informatics, University of California San Diego, La Jolla, CA, USA;Ohio State University, Columbus, OH, USA;University of Arizona, Tucson, AZ, USA;

关键词: Asthma;    Gene Expression Omnibus;    Fluticasone Propionate;    Asthma Diagnosis;    Gene Expression Omnibus Database;   

DOI  :  10.1186/1471-2105-11-S9-S8

来源: Springer

【 摘 要 】

BackgroundThe amount of data deposited in the Gene Expression Omnibus (GEO) has expanded significantly. It is important to ensure that these data are properly annotated with clinical data and descriptions of experimental conditions so that they can be useful for future analysis. This study assesses the adequacy of documented asthma markers in GEO. Three objective measures (coverage, consistency and association) were used for evaluation of annotations contained in 17 asthma studies.ResultsThere were 918 asthma samples with 20,640 annotated markers. Of these markers, only 10,419 had documented values (50% coverage). In one study carefully examined for consistency, there were discrepancies in drug name usage, with brand name and generic name used in different sections to refer to the same drug. Annotated markers showed adequate association with other relevant variables (i.e. the use of medication only when its corresponding disease state was present).ConclusionsThere is inadequate variable coverage within GEO and usage of terms lacks consistency. Association between relevant variables, however, was adequate.

【 授权许可】

CC BY   
© Lacson et al; licensee BioMed Central Ltd. 2010