Premium
SNP databases and pharmacogenetics: great start, but a long way to go
Author(s) -
Marsh Sharon,
Kwok Pui,
McLeod Howard L.
Publication year - 2002
Publication title -
human mutation
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.981
H-Index - 162
eISSN - 1098-1004
pISSN - 1059-7794
DOI - 10.1002/humu.10115
Subject(s) - false positive paradox , biology , single nucleotide polymorphism , database , snp , pharmacogenetics , hum , genetics , gene , genotype , computer science , machine learning , art , performance art , art history
With the recent publication of the human genome project there has been an explosion of data available for pharmacogenetic research. Web‐based databases containing information on single nucleotide polymorphisms (SNPs) are readily accessible to researchers, but there has been little comment on their utility. We used seven major international databases to identify SNPs in 74 genes involved in drug pathways. Very little overlap was seen among the databases, with only eight out of a putative 893 SNPs (∼1%) common to the most commonly used databases. Problems with false positives, secondary to a high degree of homology in gene families, were also observed. These studies suggest researchers limiting their studies to one database would miss a great deal of information. Effort to update compilation databases, such as HGVbase, GeneSNP, PharmGKB, and HOWDY, and the aggressive removal of false positives from all databases is required if these resources are to facilitate the intended growth in pharmacogenetics research. Hum Mutat 20:174–179, 2002. © 2002 Wiley‐Liss, Inc.