PharmGKB and the International Warfarin Pharmacogenetics Consortium: the changing role for pharmacogenomic databases and single-drug pharmacogenetics.
ABSTRACT PharmGKB, the pharmacogenetics and pharmacogenomics knowledge base (www.pharmgkb.org) is a publicly available online resource dedicated to the dissemination of how genetic variation leads to variation in drug responses. The goals of PharmGKB are to describe relationships between genes, drugs, and diseases, and to generate knowledge to catalyze pharmacogenetic and pharmacogenomic research. PharmGKB delivers knowledge in the form of curated literature annotations, drug pathway diagrams, and very important pharmacogene (VIP) summaries. Recently, PharmGKB has embraced a new role--broker of pharmacogenomic data for data sharing consortia. In particular, we have helped create the International Warfarin Pharmacogenetics Consortium (IWPC), which is devoted to pooling genotype and phenotype data relevant to the anticoagulant warfarin. PharmGKB has embraced the challenge of continuing to maintain its original mission while taking an active role in the formation of pharmacogenetic consortia.
- SourceAvailable from: PubMed Central[Show abstract] [Hide abstract]
ABSTRACT: Years of sequence feature curation by UniProtKB/Swiss-Prot, PIR-PSD, NCBI-CDD, RefSeq and other database biocurators has led to a rich repository of information on functional sites of genes and proteins. This information along with variation-related annotation can be used to scan human short sequence reads from next-generation sequencing (NGS) pipelines for presence of non-synonymous single-nucleotide variations (nsSNVs) that affect functional sites. This and similar workflows are becoming more important because thousands of NGS data sets are being made available through projects such as The Cancer Genome Atlas (TCGA), and researchers want to evaluate their biomarkers in genomic data. BioMuta, an integrated sequence feature database, provides a framework for automated and manual curation and integration of cancer-related sequence features so that they can be used in NGS analysis pipelines. Sequence feature information in BioMuta is collected from the Catalogue of Somatic Mutations in Cancer (COSMIC), ClinVar, UniProtKB and through biocuration of information available from publications. Additionally, nsSNVs identified through automated analysis of NGS data from TCGA are also included in the database. Because of the petabytes of data and information present in NGS primary repositories, a platform HIVE (High-performance Integrated Virtual Environment) for storing, analyzing, computing and curating NGS data and associated metadata has been developed. Using HIVE, 31 979 nsSNVs were identified in TCGA-derived NGS data from breast cancer patients. All variations identified through this process are stored in a Curated Short Read archive, and the nsSNVs from the tumor samples are included in BioMuta. Currently, BioMuta has 26 cancer types with 13 896 small-scale and 308 986 large-scale study-derived variations. Integration of variation data allows identifications of novel or common nsSNVs that can be prioritized in validation studies. Database URL: BioMuta: http://hive.biochemistry.gwu.edu/tools/biomuta/index.php; CSR: http://hive.biochemistry.gwu.edu/dna.cgi?cmd=csr; HIVE: http://hive.biochemistry.gwu.edu.Database The Journal of Biological Databases and Curation 02/2014; 2014:bau022. · 4.46 Impact Factor
- [Show abstract] [Hide abstract]
ABSTRACT: Abstract The Human Microbiome Project (HMP) is a global initiative undertaken to identify and characterize the collection of human-associated microorganisms at multiple anatomic sites (skin, mouth, nose, colon, vagina), and to determine how intra-individual and inter-individual alterations in the microbiome influence human health, immunity, and different disease states. In this review article, we summarize the key findings and applications of the HMP that may impact pharmacology and personalized therapeutics. We propose a microbiome cloud model, reflecting the temporal and spatial uncertainty of defining an individual's microbiome composition, with examples of how intra-individual variations (such as age and mode of delivery) shape the microbiome structure. Additionally, we discuss how this microbiome cloud concept explains the difficulty to define a core human microbiome and to classify individuals according to their biome types. Detailed examples are presented on microbiome changes related to colorectal cancer, antibiotic administration, and pharmacomicrobiomics, or drug-microbiome interactions, highlighting how an improved understanding of the human microbiome, and alterations thereof, may lead to the development of novel therapeutic agents, the modification of antibiotic policies and implementation, and improved health outcomes. Finally, the prospects of a collaborative computational microbiome research initiative in Africa are discussed.Omics: a journal of integrative biology 05/2014; · 2.29 Impact Factor
Dataset: Gupta et al,2012
HUMAN MUTATION 29(4),456^460,2008
PharmGKB and the International Warfarin
Pharmacogenetics Consortium: The Changing Role
for Pharmacogenomic Databases and Single-Drug
Ryan P . Owen, Russ B. Altman, and Teri E. Klein?
Department of Genetics, Stanford University Medical Center, Stanford, California
For the Focus Section on Pharmacogenetics
PharmGKB, the pharmacogenetics and pharmacogenomics knowledge base (www.pharmgkb.org) is a publicly
available online resource dedicated to the dissemination of how genetic variation leads to variation in drug
responses. The goals of PharmGKB are to describe relationships between genes, drugs, and diseases, and to
generate knowledge to catalyze pharmacogenetic and pharmacogenomic research. PharmGKB delivers
knowledge in the form of curated literature annotations, drug pathway diagrams, and very important
pharmacogene (VIP) summaries. Recently, PharmGKB has embraced a new role—broker of pharmacogenomic
data for data sharing consortia. In particular, we have helped create the International Warfarin
Pharmacogenetics Consortium (IWPC), which is devoted to pooling genotype and phenotype data relevant
to the anticoagulant warfarin. PharmGKB has embraced the challenge of continuing to maintain its original
mission while taking an active role in the formation of pharmacogenetic consortia. Hum Mutat 29(4), 456–460,
rrrr2008 Wiley-Liss, Inc.
KEY WORDS: warfarin; CYP2C9; VKORC1; pharmacogenetics; pharmacogenomics; PharmGKB; SNPs; IWPC
For decades, scientists have known that an individual’s
genotype can contribute to a particular phenotype, as in the case
of many hereditary diseases. In 1959, Friedrich Vogel 
defined the term pharmacogenetics as ‘‘the role of genetics in drug
responses’’—thus introducing the concept that an individual’s
DNA influences drug response. The completion of the human
genome project, coupled with accurate and cost-effective
genotyping, the emergence of high-throughput phenotyping
technologies such as microarray expression analysis, and the
discovery of important variations in the human genome have
moved the field from pharmacogenetics (one drug, one gene)
to pharmacogenomics (one drug, many genes). The promise
of genome-informed use of drugs has captured the imagination
of scientists and lay people alike.
Despite these advances, pharmacogenomics faces important
challenges to overcome before it can be routinely used in a clinical
setting. Most genome variation in humans is likely to be
functionally inconsequential [Nei, 2007], so single-gene individual
variants and their associated haplotypes must be phenotypically
characterized. Additionally, in order for pharmacogenomics to gain
widespread clinical use, we must find solutions to practical
problems such as reimbursement for genotyping and delivery of
results in a timely manner, and in a form that is interpretable by a
practicing physician. Personalized medicine also raises ethical,
legal, and moral issues with regard to patient privacy and insurance
coverage. Despite these formidable challenges, the biggest hurdle
remains a lack of basic scientific understanding about the
fundamental gene, drug, and disease relationships that contribute
to the complex phenotype of drug response and drug resistance.
PharmGKB [Klein et al., 2001] is the leading pharmacogenetics
and pharmacogenomics knowledge base. Other pharmacogenomic
databases include PharmGED (http://bidd.cz3.nus.edu.sg/phg) and
disease-specific database such as the HIV-Pharmacogenomics
database (www.hiv-pharmacogenomics.org). From its inception
until earlier this year, PharmGKB has focused primarily on the
annotation, integration, and aggregation of pharmacogenetic/
pharmacogenomic knowledge using a variety of different data
types. This mission is described in detail in previous publications
[Owen et al., 2007; Hodge et al., 2007]. Very recently, PharmGKB
has expanded on this original mission by taking a leadership role as
the broker of pharmacogenomic consortia. The first such effort in
Consortium (IWPC). PharmGKB, in collaboration with the
international community, has aggregated a genotype-phenotype
Published online 10 March 2008 in Wiley InterScience (www.
Received 24 July 2007; accepted revised manuscript 9 December
?Correspondence to:Teri E. Klein, Department of Genetics, Stan-
ford University,300 Pasteur Driver, Stanford,CA 94305.
Grant sponsors: National Institutes of Health (NIH)/National Insti-
tute of GeneralMedicalSciences
(NIGMS); Grant number:
rrrr2008 WILEY-LISS, INC.
data set of more than 5,500 subjects for the initial purpose of
developing a dosing algorithm across multiple races and ethni-
cities. We summarize these activities in order to demonstrate the
rich opportunities for pharmacogenomics of other drugs. Below we
provide some background on warfarin, and then describe
PharmGKB’s role in the warfarin story, including both our
contributions in the form of our novel pathways and very
important pharmacogene (VIP) gene summaries, as well as our
role in the formation of the IWPC www.pharmgkb.org/views/
Warfarin, a widely prescribed anticoagulant, has a narrow
therapeutic range and is difficult to dose appropriately [Rettie and
Tai, 2006]. Moreover, patients on warfarin exhibit a large degree of
interindividual variation in response [D’Andrea et al., 2005;
Kamali, 2006; Rieder et al., 2005]; overdosing of warfarin is
common and can lead to dangerous side effects, such as internal
hemorrhage. It has long been suspected that some of the variability
observed with warfarin dosing has roots in genetic variability. In
1999, a haplotype of the metabolic enzyme CYP2C9 (MIM]
601130) was found to have an influence in warfarin response
[Aithal et al., 1999], and in 2004, the vitamin K–reducing enzyme
VKORC1 (MIM] 608547) was shown to be the therapeutic target
of warfarin [Li et al., 2004; Rost et al., 2004], and also to be of
pharmacogenomic significance [Geisen et al., 2005; Rettie et al.,
2006; Rieder et al., 2005]. PharmGKB has warfarin pharmacoki-
netic (PK) and pharmacodynamic (PD) pathways, which contain
the VIP genes CYP2C9 and VKORC1 respectively, as well as
several other genes that have not yet been shown clinically to
influence warfarin response. Recently, a genetic variant of CYP4F2
was also shown to influence warfarin dose [Caldwell et al., in
Pathways for PK and PD
PharmGKB contains PK and PD pathways, both of which are
potentially important for pharmacogenetics. Each of the genes in
the PK and PD pathways for a given drug are of potential
pharmacogenetic relevance. The PharmGKB pathways can there-
fore serve as the basis for hypothesis generation, as each gene in a
specific PharmGKB pathway is a candidate gene for a pharmaco-
genetic study of that drug.
Warfarin PK and PD Pathways
Warfarin is administered as a racemic mixture of its R and S
enantiomers. The warfarin PK pathway, (www.pharmgkb.org/
search/pathway/warfarin/warfarin-pk.jsp) (Fig. 1), traces the dif-
ferent metabolic fates of the two enantiomers. S-warfarin is the
more potent form of warfarin, as indicated on the pathway. The
primary route of metabolism for S-warfarin is conversion to
7-hydroxywarfarin via CYP2C9; which is also indicated.
PharmGKB also contains the warfarin PD pathway (www.
pharmgkb.org/search/pathway/warfarin/warfarin-pd.jsp) (Fig. 2).
The two enantiomers of warfarin are both able to inhibit
VKORC1, although with different potencies. The pathway graphic
indicates that this is the primary route of inhibition. VKORC1
catalyzes the conversion of oxidized Vitamin K to reduced Vitamin
K. Treatment with warfarin blocks this reaction, which leads to a
reduction in the pool of reduced Vitamin K that is available to
serve as a cofactor for clotting proteins. Although the target of
warfarin is VKORC1, the downstream effects of the pathway are
also included. Several genes involved in clotting are also depicted.
All of these genes are downstream target genes influenced by the
form of Vitamin K that is available.
Genetic variation in CYP2C9 and VKORC1 has been identified
as playing a major role in warfarin response. CYP2C9 is involved in
the hydroxylation of S-warfarin, and VKORC1 is involved in the
conversion of oxidized to reduced Vitamin K. Given their
importance in warfarin response, VIP gene summaries have been
generated for both CYP2C9 and VKORC1.
In contrast to pathways that capture several potentially relevant
genes in a ‘‘systems’’ view, the VIP summaries focus on a single
gene in detail. Each VIP contains a detailed summary about the
gene including the key literature associated with it, drugs that are
known to interact with its protein product, and any PharmGKB
pathways that include the gene. For each VIP gene, the summary
provides at least one variant, haplotype, or splice variant.
The CYP2C9 VIP (www.pharmgkb.org/search/annotatedGene/
cyp2c9/index.jsp), contains a gene summary of CYP2C9 and its
most important variants and haplotypes. CYP2C9 is a Phase I
enzyme that metabolizes many drugs as substrates, including
warfarin. Throughout the summary text, there are links to specific
research articles that support the claim being made in the
sentence. This enables the user to specifically learn about different
portions of the VIP summary quickly, without having to click
through all of the Key PubMed ID numbers that are listed below
The VIP sections below the summary text provide a brief
reference guide. These sections provide a large amount of
information in a concise format. The user can quickly see how
many PharmGKB pathways the gene is involved in, which drugs it
interacts with, which diseases are related to the gene, and what
the important variants and haplotypes are for that gene.
Two important variants of CYP2C9, g.15450573C4T (Gen-
Bank accession number NT_030059), protein change p.Arg144Cys
pharmgkb.org/search/annotatedGene/cyp2c9/variant.jsp), define the
CYP2C9 haplotypes CYP2C9?2 and CYP2C9?3 (www.pharmgkb.
org/search/annotatedGene/cyp2c9/haplotype.jsp), respectively. Both
of these variants and their associated haplotypes have been
extensively studied in different populations, and allele frequency
tables are also included. For each variant, the position of the SNP is
unambiguously identified with a position and accession number for
genomic DNA, mRNA, and protein. Also included are the dbSNP
rs] and the UCSC Golden Path number. To our knowledge, a
PharmGKB VIP gene is the only resource that aggregates this
The VKORC1 VIP can be found at www.pharmgkb.org/search/
annotatedGene/vkorc1/index.jsp. VKORC1 catalyzes the conver-
sion of oxidized Vitamin K to reduced Vitamin K. VKORC1 is the
target gene product for warfarin and other anticoagulants. The
summary page is structured similarly to the CYP2C9 VIP , and the
important variants and haplotypes are indicated at the bottom of
the summary. Three important VKORC1 variants and four
VKORC1 haplotypes are listed.
The VKORC1 variant pages contain allele frequency tables.
Two of the variants, VKORC1g.3673G4A (GenBank AY587020)
and VKORC1g.6484C4T, are in tight linkage disequilibrium, and
HUMAN MUTATION 29(4),456^460,2008 457
the VKORC1g.6484C4T variant is often genotyped as a tagging
SNP for the causative SNP VKORC1g.3673G4A. Both of these
variants are associated with a low dose requirement for warfarin.
The third variant, VKORC1g.9041 G4A, is associated with a
There are two separate naming and numbering conventions in
the literature pertaining to VKORC1 variants, and these are
highlighted in the variant summary, and in the variant and
haplotype pages. The PharmGKB VIP variant pages provide an
appropriate forum to clarify controversies, contradictions and
confusions, which are regrettably common in pharmacogenetics.
As interest in warfarin pharmacogenetics has increased, several
warfarin research groups saw an opportunity: although each
individual group could generate a warfarin dosing algorithm that
would apply to the specific population they studied, these warfarin
dosing algorithms gave poor results when applied to other
populations. The need for a larger population (and the idea of
pooling data from multiple studies) became very attractive.
PharmGKB staff, in conjunction with Dr. Michael Caldwell
(Marshfield Clinic, Marshfield, WI) and Dr. Julie Johnson
(University of Florida, Gainesville, FL) laid the foundation for
an international warfarin consortium for the purposes of sharing
data and developing a dosing algorithm. Together, we developed a
memorandum of understanding (MOU) for the consortium
requiring that each group agree to 1) provide data for curation
and aggregation; 2) promise not to publish any independent
analyses of the data; and 3) agree to allow PharmGKB to release
the entire data set upon publication of the first paper based on the
data. The goal of this analysis is to create an improved, generally
applicable, dosing equation for warfarin. In exchange, PharmGKB
FIGURE 1. The PK pathway of warfarin. PharmGKB has standardized icons to depict genes, drugs, and metabolites. Genes or gene
products are shown as ovals, the drugs are shown as rectangles, and the metabolites are also rectangles, but with altered shading.
The S-enantiomer of warfarin is indicated as the more potent form of warfarin by the star on the upper right handcornerof the icon.
Theprimary routeofwarfarinmetabolismisindicatedby thelarger font thickness of the arrow fromS-warfarin to 7-hydroxywarfarin
via CYP2C9. Metabolites of warfarin can be visually distinguished from the parent drug by the lighter shading on the left portion of
the box.Uncertainty orlackofevidence about acertainpathway step can beindicated by a dashed arrow, such as theoneindicating
potentialeliminationofwarfarinintothebileviaABCB1. [Color ¢gurecanbeviewedintheonlineissue,whichisavailableathttp://
458HUMAN MUTATION 29(4),456^460,2008
would devote resources to bring the disparate data sets into a
common and comparable format. As a database devoted to
pharmacogenetics data sharing, PharmGKB was a natural choice
as a neutral party for data deposits and data dissemination.
Pharmacogenomics will someday allow genome-informed in-
dividualized drug therapy. It is not clear when this will occur and
what the clinical impact will be. In order to fully understand the
pharmacogenomics of a drug, one must understand: 1) the genes it
interacts with; 2) the genomic variability of those genes; and, 3)
the effect the variation has (via the drug’s PK and PD pathways)
on the efficacy and toxicity of the drug. PharmGKB pathways
attempt to help answer the first question. The PK and PD
pathways can describe the entire life cycle of the drug, from
administration, to effect, to elimination. Each gene that appears in
a pathway is of potential significance, and therefore, our
PharmGKB pathways can serve as hypothesis generating tools.
In the case of warfarin, both CYP2C9 (warfarin PK) and VKORC1
(warfarin PD) appear in the warfarin pathways, and both have
been shown to have an influence over warfarin response.
For the second question, our gene pages show a summary of the
known variation in each pharmacogene. In addition, for some
genes, we have VIP gene pages that characterize the individual
variants and haplotypes of each VIP gene that have been shown to
be of pharmacogenetic relevance.
For the third question, we have the annotated VIP variant
summaries and the primary phenotype data in the database. VIP
pages contain hand-curated information which is distilled to
provide the most essential facts and data relevant to pharmaco-
genomics. The most important papers for the VIP summaries are
provided as ‘‘Key PMIDs’’ for further reading. The VIP variant
pages also attempt to sort out nomenclature and numbering issues.
Complete variant mapping unambiguously identifies the variant in
PharmGKB adds pathways and VIP gene summaries continu-
ously, but must also maintain existing ones.
PharmGKB’s role as a broker of the IWPC was a natural
outgrowth of the daily activities of PharmGKB. Our staff was in
contact with many of the participating groups, and yet we are not
FIGURE 2. ThePDpathwayofwarfarin.ThewarfarinPDpathwayshowsthattheS-enantiomerofwarfarinismorepotent atinhibiting
reducedVitaminK. Biologicalintermediatesaredepictedwiththeboxes. Downstreamtargetgenesthataredependentontheformof
VitaminK that is available for their role inclotting and other physiological processes are also shown. [Color ¢gurecan beviewed in
theonlineissue,whichis available at http://www.interscience.wiley.com.]
HUMAN MUTATION 29(4),456^460,2008 459
direct scientific competitors with any of them. The emergence of
the warfarin consortium may reflect a shift in paradigms away from
a single gene–dominated view of pharmacogenetics, and toward a
more routine consideration of multigenic effects. Under this
model, genetic variation in several genes can all make small
contributions that together, produce a complex phenotype. The
small effects, multigene model would likely lead to an increase in
whole-genome type studies, which is arguably already happening
in the literature, as well as an increased role for databases such as
PharmGKB to collect information in a manner that will catalyze
research. To assess these models, a large subject population is
required. If this is indeed the future of pharmacogenomics, then it
is extremely likely that other consortia will need to be formed for
drugs with known adverse effects. An increase in consortia will
result in an evolving role for pharmacogenomics databases such as
PharmGKB, as these databases must maintain their original
purpose to describe the variation in genes across a wide variety of
different drugs while having the flexibility to engage in an in-depth
analysis of single drug pharmacogenomics. Databases will therefore
have to take on a more active role in primary research, becoming a
participant as well as a repository if future consortia are to be
successful. A trusted and experienced third party intermediary
with an interest in pharmacogenomics data sharing is very useful
in these endeavors. The shift from a role as a data repository to a
dual role as a traditional database as well as a partner in the
generation of drug consortia may also have implications for
resource allocation of databases away from the development side
and towards the curation side, as experienced biologists will be
needed at databases to work with the primary scientific submitters.
Preliminary efforts are already underway at PharmGKB for the
formation of other consortia.
The possibility that important drug-response phenotypes result
from many small contributions underscores the need for complete
pathways, characterization of functional variants, and data sharing
and communication infrastructure.
PharmGKB is supported by NIH/NIGMS grant UO1GM61374.
Aithal GP , Day CP , Kesteven PJ, Daly AK. 1999. Association of
polymorphisms in the cytochrome P450 CYP2C9 with warfarin dose
requirement and risk of bleeding complications. Lancet 353:717–719.
Caldwell MD, Awad T, Johnson JA, Gage BF, Falkowski M, Gardina P ,
Hubbard J, Turpaz Y, Langaee TY, Eby C, King C, Brower A, Schmelzer
JR, Glurich I, Vidaillet HJ, Yale SH, Zhang KQ, Berg RL, Burmester JK.
2008. CYP4F2 genetic variant alters required warfarin dose. Blood
D’Andrea G, D’Ambrosio RL, Di Perna P , Chetta M, Santacroce R,
Brancaccio V, Grandone E, Margaglione M. 2005. A polymorphism in
the VKORC1 gene is associated with an interindividual variability in the
dose-anticoagulant effect of warfarin. Blood 105:645–649.
Geisen C, Watzka M, Sittinger K, Steffens M, Daugela L, SeilFried E,
Mu ¨ller CR, Wienker TF, Oldenburg J. 2005. VKORC1 haplotypes and
their impact on the inter-individual and inter-ethnical variability of oral
anticoagulation. Thromb Haemost 94:773–779.
Hodge AE, Altman RB, Klein TE. 2007. The PharmGKB: integration,
aggregation, and annotation of pharmacogenomic data and knowledge.
Clin Pharmacol Ther 81:21–24.
Kamali F. 2006. Genetic influences on the response to warfarin. Curr Opin
Klein TE, Chang JT, Cho MK, Easton KL, Fergerson R, Hewett M, Lin Z,
Liu Y, Liu S, Oliver DE, Rubin DL, Shafa F, Stuart JM, Altman RB.
2001. Integrating genotype and phenotype information: an overview of
the PharmGKB project. Pharmacogenetics Research Network and
Knowledge Base. Pharmacogenomics J 1:167–170.
Li T, Chang CY, Jin DY, Lin PJ, Khvorova A, Stafford DW. 2004.
Identification of the gene for vitamin K epoxide reductase. Nature
Nei M. 2007. The new mutation theory of phenotypic evolution. Proc Natl
Acad Sci USA 104:12235–12242.
Owen RP , Klein TE, Altman RB. 2007. The Education potential of the
(PharmGKB). Clin Pharmacol Ther 82:472–475.
Rettie AE, Tai G. 2006. The pharmocogenomics of warfarin: closing in on
personalized medicine. Mol Interv 6:223–227.
Rettie AE, Farin FM, Beri NG, Srinouanprachanh SL, Rieder MJ,
Thijssen HH. 2006. A case study of acenocoumarol sensitivity and
polymorphisms in VKORC1 and CYP2C9. Br J Clin Pharmacol
Rieder MJ, Reiner AP , Gage BF, Nickerson DA, Eby CS, McLeod HL,
Blough DK, Thummel KE, Veenstra DL, Rettie AE. 2005. Effect of
VKORC1 haplotypes on transcriptional regulation and warfarin dose. N
Engl J Med 352:2285–2293.
Rost S, Fregin A, Ivaskevicius V, Conzelmann E, Ho ¨rtnagel K, Pelz HJ,
Lappegard K, Seifried E, Scharrer I, Tuddenham EG, Mu ¨ller CR, Strom
TM, Oldenburg J. 2004. Mutations in VKORC1 cause warfarin
resistance and multiple coagulation factor deficiency type 2. Nature
Vogel F. 1959. Moderne Probleme der Humangenetik. Ergeb Inn Med
explainedby combinations of
460HUMAN MUTATION 29(4),456^460,2008