A gene by any other name doesn't smell as sweet: Identifying and correcting mislabeled gene names in genomic databases

Human genes are identified using various naming conventions; the most universal among biological and public health literature is the gene symbol.  Gene symbols are the nomenclature of choice because they are easy to read, remember and often abbreviate information about the gene function.  A list of valid and current gene symbols is maintained by The Human Gene Nomenclature Committee for the sake of uniformity. Gene symbols, however, are prone to error. Genes are sometimes referred to by multiple aliases in the literature, and can be unintentionally converted into dates by common spreadsheet programs. Outdated and invalid gene symbols can accumulate in long-term databases such as the Gene Expression Omnibus (GEO) of the National Center for Biotechnology Information. In an effort to quantify the extent of the mislabelled and out of date gene symbols housed in public databases, a map of known gene symbol errors was created and combined with the HGNC database through the R package, HGNChelper.  Analysis of annotations for all Homo Sapiens genomic platforms in GEO revealed up to 15% invalid or out of date gene symbols for some assays. The suggested method is a quick and simple solution to a problem affecting any researcher working with gene symbols.