- Atlas
- Medicine & Health
- Pharmacogenomics
Pharmacogenomics
Pharmacogenomics is the study of how a person's genes affect their response to drugs.
8 frameworks, from 1957 to today · Pharmacology · Medicine & Health
- Classical Pharmacogenetics 1957-1995
- Metabolizer Phenotype Classification 1970-Present
- Pharmacogenomics and Genome-Wide Discovery 1995-Present
- Biomarker and Companion Diagnostic Model 1998-Present
- Pharmacogenomics of oncology and targeted therapy 2000-Present
- Guideline Translation and Clinical Implementation 2009-Present
- Preemptive Panel Testing 2010-Present
- Ancestry-Aware Pharmacogenomics 2015-Present
Overview
Pharmacogenomics is the study of how a person's genes affect their response to drugs1. The field rests on a simple but consequential observation: the same dose of the same medicine can cure one patient, do nothing for another, and poison a third. Its practitioners ask which inherited differences explain that variation, how those differences work at the level of enzymes, transporters, and receptors, and how to turn the answer into a test that guides prescribing1. The stakes are practical—safer doses, fewer adverse reactions, and drugs matched to the biology of a tumor rather than to a diagnostic label. The history below follows the field from the first hints that heredity shapes drug response to the present era of panel testing and ancestry-aware guidelines.
Heredity and drug response before the gene
Long before anyone knew what a gene was, clinicians noticed that drug reactions ran in families. The modern field usually dates its founding to a short 1957 paper by the German pharmacologist Friedrich Vogel, who coined the term pharmacogenetics for the study of inherited differences in drug response. But the observations came earlier. In the 1930s, clinicians reported that some patients given the antimalarial primaquine developed severe hemolysis—destruction of red blood cells—while others did not, and that the sensitivity appeared familial. During the Second World War, military physicians noted that some soldiers given the antimalarial drug pamaquine suffered the same reaction. The explanation came in the 1950s, when researchers linked the hemolysis to a deficiency of the enzyme glucose-6-phosphate dehydrogenase (G6PD), a hereditary trait common in some populations. This was an early, concrete demonstration that a single inherited enzyme defect could turn a standard drug into a hazard1.
A second landmark came from anesthesia. In the 1950s, reports accumulated of patients who, after receiving the muscle relaxant succinylcholine during surgery, remained paralyzed for hours instead of minutes1. The cause was eventually traced to inherited variants of the enzyme butyrylcholinesterase, which normally breaks the drug down1. People with low-activity forms of the enzyme cleared the drug slowly and stayed paralyzed longer. These cases established the field's core logic: a drug's fate in the body depends on enzymes, and enzymes vary between people because of heredity.
The rise of pharmacogenetics as a science
Through the 1960s and 1970s, pharmacogenetics grew into a recognizable research program1. Investigators studied twins to separate genetic from environmental contributions to drug handling, and they mapped the inheritance of traits such as the acetylator status—the speed at which a person's liver acetylates certain drugs, governed by the enzyme N-acetyltransferase. The discovery in the 1970s that the anticonvulsant phenytoin and other drugs were metabolized by a family of liver enzymes, the cytochrome P450 system, gave the field its central cast of characters. Debrisoquine, an antihypertensive, became a classic probe: some people metabolized it rapidly and others poorly, and the poor metabolizers were later shown to carry variants in the gene CYP2D6. By the 1980s, the field could describe, for a growing list of drugs, which enzyme handled them and which inherited variants changed the handling.
The crucial conceptual step was the classification of genetic polymorphism—the existence of common variants in a population—into phenotypes. A person could be a poor, intermediate, extensive, or ultrarapid metabolizer of a given enzyme, and that status predicted drug levels. This phenotype language, still in use, made pharmacogenetics clinically legible: it translated a genotype into a prediction about how a patient would handle a drug. The field also learned that the same variant could matter differently depending on the drug and the enzyme's role—a poor metabolizer of a prodrug that must be activated might get no benefit, while a poor metabolizer of a drug that is inactivated might accumulate toxic levels.
From single genes to the genome
The term pharmacogenomics emerged in the 1990s, as the Human Genome Project and new sequencing technologies made it possible to study many genes at once rather than one enzyme at a time. Where pharmacogenetics had focused on single-gene traits, pharmacogenomics promised a genome-wide view: the variants that shape drug response across the whole of a person's DNA. The change was partly rhetorical and partly real. It brought the field into contact with genomics, bioinformatics, and the pharmaceutical industry, which saw in it a way to explain why drugs failed in trials and to rescue compounds that worked only in a subset of patients.
The 1990s and 2000s produced the discoveries that anchor clinical practice today. The enzyme thiopurine methyltransferase (TPMT) was shown to determine toxicity of thiopurine drugs used in leukemia and autoimmune disease; patients with low-activity TPMT accumulated toxic metabolites. Variants in CYP2C19 were linked to the activation of clopidogrel, an antiplatelet drug that is a prodrug, so that poor metabolizers got less protection against clots12. The HLA gene HLA-B\57:01 was found to predict a severe hypersensitivity reaction to the HIV drug abacavir, and HLA-B\15:02 to predict a dangerous skin reaction to the anticonvulsant carbamazepine in some populations. These were not merely associations; they were actionable predictions, and they set the stage for the biomarker era.
Biomarkers, companion diagnostics, and oncology
A pharmacogenomic biomarker is a measurable genetic feature that predicts drug response, and a companion diagnostic is a test marketed alongside a drug to identify the patients who should receive it3. The oncology field drove this model hardest. The breast cancer drug trastuzumab, approved in the late 1990s, was effective only in tumors that overexpressed the HER2 receptor, and testing for HER2 became standard. Later, targeted therapies were paired with tests for mutations in genes such as EGFR, BRAF, and KRAS, so that a lung or melanoma tumor's genotype, not just its tissue of origin, determined the drug. This was the "magic bullet" idea revived at the molecular level: a drug aimed at the specific abnormality driving a cancer.
The oncology model also exposed the field's central tension. Tumor genomics is somatic—mutations acquired by the cancer—while inherited pharmacogenomics concerns the patient's own germline. Both matter, and both require tests, but they answer different questions: whether a drug will hit its target, and whether the patient will handle the drug safely. The companion-diagnostic framework, built for the first question, had to be extended to the second.
Clinical implementation and guidelines
By the 2010s, the field's problem was no longer discovery but translation: how to get a genotype into a prescribing decision at the point of care. Professional consortia took up the task. The Clinical Pharmacogenetics Implementation Consortium (CPIC), formed in 2009, began publishing peer-reviewed guidelines that translate genotype into dosing recommendations for specific drug–gene pairs. The Dutch Pharmacogenetics Working Group produced similar guidance. These bodies did not decide whether to test; they assumed a genotype was available and told clinicians what to do with it1. That framing was deliberate, because the evidence for testing everyone was still thin, while the evidence for acting on a known genotype was stronger.
Implementation also required infrastructure: electronic health records that could store genotypes and fire alerts, laboratory standards for reporting, and reimbursement. Early institutional programs—at St. Jude Children's Research Hospital, which built pharmacogenomic testing into pediatric cancer care, and at several academic medical centers—demonstrated that the workflow could function. The field learned that the barrier was rarely the science; it was the plumbing.
Preemptive panel testing and population-aware practice
The most recent turn is preemptive panel testing: testing a patient for many pharmacogenes at once, before any specific drug is prescribed, so that the result is on file when a prescription is written. This inverts the older model of testing reactively, when a drug is already chosen. Panels covering CYP2D6, CYP2C19, CYP2C9, VKORC1, TPMT, and HLA genes are now offered by several health systems and commercial labs. The approach is attractive because a single test can inform many future prescriptions, but it raises questions about which variants to include, how to handle variants of uncertain significance, and how to keep results current as knowledge changes.
Running alongside this is a reckoning with ancestry. Many pharmacogene variants differ in frequency across populations—CYP2D6 ultrarapid metabolizers are more common in some African and Oceanian populations, HLA-B\15:02 is more frequent in parts of East and Southeast Asia, and CYP2C19* poor metabolizers are common in East Asia and Oceania. Guidelines increasingly warn against using self-identified race as a proxy for genotype, while acknowledging that ancestry-informed testing strategies can improve detection. The field is working toward panels and algorithms that account for population structure without reducing it to a crude category.
The present
Today pharmacogenomics is a mature translational science with a recognized set of drug–gene pairs, published guidelines, and growing clinical adoption. Its main approaches descend directly from the history above: phenotype classification from the early pharmacogenetic era, genome-wide discovery from the 1990s, biomarker-and-companion-diagnostic pairing from oncology, and guideline translation and preemptive testing from the implementation era. The open questions are practical and unresolved: how to generate the randomized evidence that convinces payers and skeptics, how to handle the many variants whose effects are unknown, how to make preemptive testing equitable across populations, and how to keep guidelines synchronized with a literature that grows faster than any clinician can read. The field's founding insight—that heredity shapes drug response—is no longer in doubt; what remains is the work of making that insight reliably useful at the bedside.
References
- Pharmacogenomics (introduction, Pharmacokinetics, Off-target sites, History, Applications, Clinical Pharmacogenetics Implementation Consortium (CPIC)), Wikipedia
- CYP2C19, Wikipedia
- Companion diagnostic, Wikipedia
The frameworks of Pharmacogenomics
Each framework is a way of working the field has used: when it held, and the question it answers.
- 1957-1995
Classical Pharmacogenetics
How do inherited differences shape drug response?
- 1970-Present
Metabolizer Phenotype Classification
How fast does a person metabolize a given drug?
- 1995-Present
Pharmacogenomics and Genome-Wide Discovery
Which genes across the genome affect drug response?
- 1998-Present
Biomarker and Companion Diagnostic Model
Which genetic test identifies patients for a drug?
- 2000-Present
Pharmacogenomics of oncology and targeted therapy
Which tumor mutations predict response to targeted drugs?
- 2009-Present
Guideline Translation and Clinical Implementation
What should a clinician do with a known genotype?
- 2010-Present
Preemptive Panel Testing
Should we test many pharmacogenes before prescribing?
- 2015-Present
Ancestry-Aware Pharmacogenomics
How do drug-response variants differ across ancestries?
How the ideas connect
Each arrow runs from an earlier idea to the later one it shaped. The list below says what passed between them.
Metabolizer Phenotype Classification grew out of Classical Pharmacogenetics
The phenotype classification of poor, intermediate, extensive and ultrarapid metabolizers grew directly out of classical pharmacogenetics' study of inherited enzyme variants such as the acetylator polymorphism and CYP2D6 debrisoquine oxidation. Investigators turned single-gene enzyme defects into a predictive vocabulary of metabolizer status.
Carried by Robert Smith; Michel Eichelbaum.
Pharmacogenomics and Genome-Wide Discovery replaced Classical Pharmacogenetics
Pharmacogenomics replaced the single-gene, single-enzyme program of classical pharmacogenetics with a genome-wide view enabled by the Human Genome Project and new sequencing methods. The older focus on one enzyme at a time became a limiting case within the broader genomic approach.
Carried by Human Genome Project; William Evans; Mary Relling.
Biomarker and Companion Diagnostic Model grew out of Pharmacogenomics and Genome-Wide Discovery
The biomarker and companion diagnostic model grew out of pharmacogenomic discovery: once variants such as HLA-B*57:01 and HER2 amplification were shown to predict drug response, they were packaged as tests marketed alongside drugs. Genome-wide association findings supplied the biomarkers that the companion diagnostic framework operationalized.
Carried by Trastuzumab; HLA-B*57:01.
Pharmacogenomics of oncology and targeted therapy grew out of Biomarker and Companion Diagnostic Model
Pharmacogenomics of oncology and targeted therapy is the branch of the biomarker/companion diagnostic model applied to tumor genomics, pairing drugs with tests for EGFR, BRAF and KRAS mutations. The companion diagnostic framework was built for and driven hardest by oncology.
Carried by Trastuzumab; Imatinib.
Pharmacogenomics and Genome-Wide Discovery influenced by Metabolizer Phenotype Classification
The metabolizer phenotype vocabulary carried over into pharmacogenomics as the interpretive frame for genome-wide variant discovery: CYP2D6 and CYP2C19 variants found by genomic methods were still reported as poor or ultrarapid metabolizer status. The phenotype language gave genomic findings clinical meaning.
Ancestry-Aware Pharmacogenomics influenced by Metabolizer Phenotype Classification
Ancestry-aware pharmacogenomics took the metabolizer phenotype categories and asked how their frequencies differ across populations, for example the higher prevalence of CYP2D6 ultrarapid metabolizers in some African and Oceanian groups. The phenotype classification supplied the categories whose population distribution ancestry analysis then explained.
Guideline Translation and Clinical Implementation influenced by Pharmacogenomics and Genome-Wide Discovery
Guideline translation took the drug-gene associations established by pharmacogenomic discovery, such as TPMT-thiopurine and CYP2C19-clopidogrel, and turned them into dosing recommendations. CPIC, formed in 2009, assumed a genotype was available and told clinicians what to do with it.
Carried by Clinical Pharmacogenetics Implementation Consortium.
Guideline Translation and Clinical Implementation influenced by Biomarker and Companion Diagnostic Model
The biomarker model's logic of acting on a known predictive genotype carried into guideline translation: CPIC guidelines were written for specific drug-gene pairs whose predictive value had been established in the biomarker era. The guidelines translated established biomarkers into prescribing rules.
Carried by Clinical Pharmacogenetics Implementation Consortium.
Preemptive Panel Testing grew out of Guideline Translation and Clinical Implementation
Preemptive panel testing grew out of guideline translation: once CPIC and Dutch guidelines specified what to do with many drug-gene pairs, it became attractive to test for many genes at once before prescribing. The guidelines supplied the interpretive framework that made a stored panel result actionable.
Carried by Clinical Pharmacogenetics Implementation Consortium.
Ancestry-Aware Pharmacogenomics influenced by Preemptive Panel Testing
Preemptive panel testing raised the question of which variants to include and how to interpret them across diverse patients, prompting ancestry-aware approaches to variant selection and frequency. The panel format created the practical setting in which ancestry-aware interpretation became necessary.
Beyond the field
Where Pharmacogenomics's ideas came from in other fields, and what they led to there: the strongest 12 of 32 links.
Classical Pharmacogenetics led to Pharmacogenomics in Translational Pharmacology
Pharmacogenomics grew directly out of classical pharmacogenetics, extending single-gene drug-response variation to genome-wide analysis. The earlier term and findings were generalised as genomic methods arrived.
Classical Pharmacogenetics led to Pharmacogenomics in Pharmacology
Classical pharmacogenetics, the study of inherited differences in drug metabolism and response from the 1950s onward, is the direct ancestor of pharmacogenomics; the later field generalised it from single-gene traits to genome-wide analysis.
Carried by Friedrich Vogel.
Classical Pharmacogenetics led to Pharmacogenomic pharmacokinetics in Pharmacokinetics
Classical pharmacogenetics, founded on inherited enzyme polymorphisms such as pseudocholinesterase and acetylation status, supplied the concepts and case material that pharmacogenomic pharmacokinetics generalised to genome-wide markers and routine dosing.
Carried by Friedrich Vogel; Werner Kalow.
Classical Pharmacogenetics led to Pharmacogenomics in Medical Genetics
Pharmacogenomics grew directly out of classical pharmacogenetics: the study of inherited variation in drug response was extended from single-gene enzyme polymorphisms to genome-wide variation as sequencing and genotyping scaled up. The questions and clinical aims of pharmacogenetics carried over into the renamed field.
Carried by Friedrich Vogel; Arno Motulsky.
Classical Pharmacogenetics led to Pharmacogenomics in Clinical Pharmacology
Classical pharmacogenetics, the study of inherited single-gene variation in drug response from the 1950s, is the direct parent of pharmacogenomics. The latter extended the same questions to genome-wide variation and drug discovery.
Carried by Friedrich Vogel; Arno Motulsky.
Pharmacogenomics and Genome-Wide Discovery influenced Personalized medicine in Clinical Decision Science
Pharmacogenomics supplied the core evidence base for personalized medicine: genotype-guided drug selection and dosing, built on the Human Genome Project and genome-wide association studies. The 2003 completion of the genome sequence and subsequent pharmacogenomic discovery are what made individualized prescribing a clinical programme rather than an aspiration.
Carried by Human Genome Project; International HapMap Project.
Biomarker and Companion Diagnostic Model led to Comprehensive Genomic Profiling in Precision Oncology
The biomarker and companion diagnostic model is the framework that comprehensive genomic profiling formalized and scaled up: companion diagnostics established that a genomic biomarker must be validated and tied to a specific drug, and profiling extended this from single biomarkers to hundreds of genes at once.
Ancestry-Aware Pharmacogenomics influenced by Statistical and Population Genomics in Bioinformatics
Statistical and population genomics supplied the methods, ancestry-informative markers, admixture mapping and population-frequency reference panels, by which pharmacogenomic variant frequencies were compared across populations. Ancestry-aware pharmacogenomics is built on these population-genetic tools.
Classical Pharmacogenetics influenced by Clinical pharmacology in Pharmacology
Clinical pharmacology, established as a discipline around 1950, supplied the methods of measuring drug levels and individual variation in patients that classical pharmacogenetics used to tie inherited enzyme variants to drug response.
Classical Pharmacogenetics influenced Candidate gene in Genetic Epidemiology
Classical pharmacogenetics established that inherited variation in single genes explains drug response, providing the model of candidate-gene association that candidate gene studies generalised to disease. The single-gene, hypothesis-driven logic of pharmacogenetics carried over.
Metabolizer Phenotype Classification influenced by Pharmacokinetics in Pharmaceutical Chemistry
Pharmacokinetic concepts of clearance and half-life, formalised in the 1960s, gave the basis for classifying people by how fast they metabolise drugs into poor, intermediate, extensive and ultrarapid phenotypes.
Pharmacogenomics and Genome-Wide Discovery influenced Personalized medicine in Clinical Medicine
Pharmacogenomics supplied personalized medicine with its central method of using genetic variation to predict drug response, and genome-wide discovery studies provided the variant data on which stratification depends.

Pharmacogenomics is a planet in Nousfarer
Its city has a tower for each of these 8 frameworks. Land, walk in, and climb one concept at a time.
Land on PharmacogenomicsRelated fields
Fields its ideas are linked to
- Precision Oncology 6 links
- Translational Pharmacology 4 links
- Pharmacokinetics 3 links
- Medical Genetics 2 links
- Clinical Medicine 2 links
- Pharmacology 2 links
- Bioinformatics 1 link
- Gastroenterology And Hepatology 1 link
- Clinical Decision Science 1 link
- Evidence Based Medicine 1 link
- Quality Improvement 1 link
- Genetic Epidemiology 1 link
More in Pharmacology
- Clinical Pharmacology
- Drug Discovery
- Pharmacodynamics
- Pharmacokinetics
- Pharmacology
- Pharmacometrics
- Pharmacovigilance
- Receptor Theory And Signal Transduction
- Toxicology
- Translational Pharmacology
See where Pharmacogenomics sits on the Map of Human Knowledge