A growing number of companies are offering a seductive proposition: spit in a tube, send your DNA to a lab, and receive a personalized cannabis recommendation based on your genetic profile. The pitch is that your genes determine how you metabolize THC, how sensitive your cannabinoid receptors are, and how likely you are to experience anxiety versus relaxation from any given strain.

The concept is rooted in pharmacogenomics — the legitimate science of how genetic variation affects drug response. And some of the underlying science is real. But the gap between what genetics can currently tell us about cannabis response and what these companies claim to offer is significant.

What Genetics Can Actually Tell Us

Several well-characterized genetic variants do influence cannabis response. The science here is genuine:

CYP2C9: The Metabolism Gene

The cytochrome P450 enzyme CYP2C9 is responsible for metabolizing THC into its primary inactive metabolite, 11-nor-9-carboxy-THC. Genetic variants in CYP2C9 produce meaningful differences in THC metabolism speed:

CYP2C9 *1/*1 (wild type): Normal THC metabolism. This is the most common genotype, present in approximately 65-70% of people of European descent.

CYP2C9 *1/*3 (heterozygous variant): Slower THC metabolism. Carriers of one copy of the *3 allele metabolize THC approximately 30% more slowly than wild-type individuals. This means the same dose produces higher peak blood levels and longer-lasting effects.

CYP2C9 *3/*3 (homozygous variant): Significantly slower THC metabolism. These individuals — approximately 1-3% of the population — metabolize THC up to three times more slowly than wild-type. They are substantially more sensitive to THC and may experience pronounced effects from doses that other people consider mild.

This is clinically meaningful. A person with the *3/*3 genotype who takes a 10mg edible may experience effects equivalent to what a *1/*1 individual would feel from a 25-30mg dose. For these individuals, the standard “start low, go slow” advice is even more important — and the appropriate starting dose may be lower than general recommendations suggest.

AKT1: The Psychosis Risk Gene

The AKT1 gene encodes a protein kinase involved in dopamine signaling in the brain. A specific variant — the C/C genotype at rs2494732 — has been associated with increased risk of psychotic experiences in response to cannabis use.

A 2012 study published in Neuropsychopharmacology found that daily cannabis users with the AKT1 C/C genotype were seven times more likely to experience psychotic symptoms compared to users with the T/T genotype. The effect was dose-dependent — occasional use carried lower risk than daily use even in genetically susceptible individuals.

This finding has been replicated in several studies and represents one of the more robust gene-environment interactions in cannabis research. It does not mean that everyone with the C/C genotype will experience psychosis — the vast majority will not. But it does identify a subpopulation for whom heavy cannabis use carries elevated psychiatric risk.

COMT: The Anxiety Gene

The catechol-O-methyltransferase (COMT) gene affects dopamine metabolism in the prefrontal cortex. The Val158Met polymorphism produces two common variants:

Val/Val: Faster dopamine breakdown, associated with lower baseline prefrontal dopamine levels. These individuals may be more susceptible to cannabis-induced anxiety because THC-driven dopamine release pushes prefrontal dopamine above optimal levels more easily.

Met/Met: Slower dopamine breakdown, associated with higher baseline prefrontal dopamine levels. Some research suggests these individuals may experience more cognitive impairment from cannabis but potentially less anxiety.

The COMT-cannabis interaction is less robust than the AKT1 finding and has produced inconsistent results across studies. It is suggestive rather than definitive.

CNR1: The Receptor Gene

The CNR1 gene encodes the CB1 cannabinoid receptor — the primary target for THC in the brain. Several polymorphisms in CNR1 have been associated with differences in cannabis sensitivity, reward processing, and susceptibility to cannabis use disorder.

The (AAT)n repeat polymorphism in CNR1 has been linked to cannabis dependence susceptibility in some studies, with longer repeat lengths associated with increased risk. Other CNR1 variants have been associated with differences in THC’s subjective effects, including variations in euphoria, anxiety, and psychomotor impairment.

FAAH: The Endocannabinoid Breakdown Gene

FAAH (fatty acid amide hydrolase) is the primary enzyme that breaks down anandamide — the body’s endogenous cannabinoid. The FAAH C385A variant produces a version of the enzyme that degrades more quickly, resulting in higher baseline anandamide levels in the brain and body.

Individuals with the A/A genotype at this position have chronically elevated anandamide levels and tend to show:

  • Lower baseline anxiety
  • Greater resilience to stress
  • Potentially different responses to external cannabinoids because their endocannabinoid system is already more active

What Genetics Cannot Tell Us

Here is where the commercial DNA testing services overreach:

Strain-specific recommendations are not genetically derivable. No genetic test can tell you that you will enjoy Blue Dream more than OG Kush. Strain effects depend on the specific combination of cannabinoids, terpenes, and minor compounds in each batch — plus the entourage effects of their interaction — in ways that current genetic science cannot predict with strain-level specificity.

Terpene sensitivity is not well-characterized genetically. While olfactory receptor genetics influence how you perceive aromas (and therefore contribute to strain preference), the connection between genetic terpene perception and subjective cannabis effects is almost entirely unknown.

Set and setting overwhelm genetics for most consumers. Your mood, your environment, your expectations, your tolerance level, your recent sleep quality, and what you ate that day all influence your cannabis experience more than genetic factors for the vast majority of people. Genetics sets a range; everything else determines where within that range you land on any given occasion.

The studies are small and often unreplicated. The gene-cannabis interaction literature is still young. Many findings come from small studies in specific populations and have not been consistently replicated. Building commercial products on this evidence base is premature.

The Commercial Landscape

Several companies now offer cannabis pharmacogenomic testing:

The typical workflow involves providing a saliva or cheek swab DNA sample, which is genotyped for a panel of relevant variants (usually including CYP2C9, AKT1, COMT, CNR1, and FAAH among others). Results are delivered as a report that may include:

  • THC metabolism speed estimate
  • Relative risk assessment for anxiety or psychotic experiences
  • General dosing guidance (conservative vs. standard)
  • Product type recommendations (flower vs. edibles vs. topicals)
  • Strain archetype suggestions

The useful parts of these reports — CYP2C9 metabolism classification and AKT1 risk assessment — are genuinely informative. A person who discovers they are a slow THC metabolizer has actionable information that could prevent uncomfortable overconsumption experiences. A person who carries the AKT1 C/C genotype has information relevant to their decisions about cannabis use frequency and intensity.

The strain-matching and product-matching components are speculative. They may be directionally interesting — a slow metabolizer might indeed do better with lower-potency products — but the specificity of recommendations typically exceeds what the underlying genetics can support.

The Future of Personalized Cannabis

The vision of genetically personalized cannabis is not wrong — it is early. As genomic databases grow, as cannabis response phenotyping becomes more sophisticated, and as the molecular pharmacology of cannabinoids and terpenes becomes clearer, the ability to predict individual cannabis responses from genetic data will improve.

Within the next decade, it is plausible that genetic testing could provide:

  • Reliable THC dose optimization based on metabolism genetics
  • Meaningful risk stratification for adverse psychiatric effects
  • Evidence-based guidance on cannabinoid ratios likely to produce desired effects
  • Identification of individuals for whom cannabis is medically contraindicated based on genetic susceptibility

Today, cannabis genetic testing offers some useful information wrapped in significant overstatement. The metabolism and risk genes are worth knowing about. The strain-matching is marketing.

As with most things in cannabis: the science is real, the hype outruns it, and the truth is interesting enough without the embellishment.