Translating Diplotypes Into Enzyme Activity

From Diplotype to Functional Input

A diplotype identifies the two inherited haplotypes at a gene, but a PBPK model needs a functional representation of what those haplotypes imply for enzyme activity. The translation is gene-specific: an allele’s functional value, the number of gene copies, and the rules used to assign a phenotype all matter. A metabolizer label is a useful summary, but it is not itself a measured clearance or a universal multiplier for enzyme abundance.

The earlier material on alleles and diplotypes establishes how star alleles describe haplotypes and how two haplotypes form a diplotype. Here, assume the diplotype has been assigned correctly. The task is to translate it into a functional category or score, while keeping the assumptions and uncertainty visible.

For CYP2D6, a commonly used activity-score convention assigns no-function alleles a value of 0, decreased-function alleles a value such as 0.25 or 0.5 according to the applicable reference, normal-function alleles a value of 1, and increased-function alleles a value greater than 1 when supported by that scoring system. The score is the sum across functional gene copies. Under commonly used phenotype cutoffs, a score of 0 corresponds to poor metabolizer (PM), a score greater than 0 and less than 1.25 to intermediate metabolizer (IM), a score from 1.25 through 2.25 to normal metabolizer (NM), and a score greater than 2.25 to ultrarapid metabolizer (UM). Use the cutoffs specified by the relevant guideline and version; boundary conventions should not be silently mixed.

CYP2C19 is often translated by diplotype-to-phenotype rules rather than by summing a CYP2D6-style activity score. For the common alleles illustrated below, *1 is normal function, *2 and *3 are no function, and *17 is increased function. The phenotype assignment depends on the pair of alleles: for example, *1/*2 is IM, *2/*2 is PM, *1/*1 is NM, *1/*17 is rapid metabolizer (RM), and *17/*17 is UM. A *2/*17 diplotype is generally classified as IM in commonly used guidance. The applicable gene-specific reference is essential, especially for less common alleles and copy-number variants.

Assumptions Behind the Translation

A numerical score is only as defensible as the information used to construct it. The following assumptions should be checked before treating a diplotype as a PBPK input:

• Allele assignment is sufficiently reliable. The star-allele call must be supported by the variant data and the assay’s ability to detect relevant alleles. If phasing or allele assignment is uncertain, the functional result may also be uncertain.
• Functional values match the gene and reference system. Do not transfer a value from one gene to another or assume every laboratory uses identical values and cutoffs.
• Copy number is resolved where it matters. A deletion or duplication can change the number of functional copies. For CYP2D6, the duplicated allele must be identified where possible; a total copy count alone may not reveal how many copies are functional.
• The score represents inherited function, not every cause of current activity. Disease, interacting drugs, and other acquired influences can change observed enzyme function. A genotype-based category does not automatically account for those influences.
• Translation to a model parameter is separately justified. A phenotype category is not automatically equal to a particular fold-change in enzyme abundance, intrinsic clearance, or whole-body clearance.

These assumptions do not imply that a score captures all biology. They specify what must be true for the chosen scoring procedure to be interpretable as a functional summary.

Why CYP2D6 Activity Scores Add Across Copies

For a CYP2D6 diplotype, let \(a_j\) be the activity value assigned to copy \(j\), with \(j\) indexing all gene copies that contribute. The activity score is

where \(m\) is the number of copies included. The sum is a scoring convention, not a claim that every copy produces identical protein in every tissue. Its rationale is that, if each copy contributes independently and its assigned value is proportional to its relative functional contribution, total relative activity is the sum of those contributions. For example, two copies each assigned 1 contribute \(1+1=2\) score units under that convention. If copy-specific expression or function is uncertain, the resulting score is an approximation rather than a direct assay result.

A deletion is represented by absence of a functional copy, not by assigning a normal-function value to a missing gene. In CYP2D6 nomenclature, *5 is a commonly used whole-gene deletion allele and contributes 0. A duplication or multiplication, often indicated by a copy-number suffix such as *1x2, means that the specified allele occurs more than once; its activity value is counted for each confirmed copy.

Deletion and Duplication Change the Copy Calculation

For copy-number variation, count the functional contributions rather than treating a diplotype as exactly two equivalent copies. The usual two-haplotype description still matters, but one haplotype may contain a deletion or multiple copies. The identity of the duplicated allele is especially important: duplicating a no-function allele does not have the same consequence as duplicating a normal-function allele.

A high activity score is a category assignment under the chosen system, not proof of a specific clinical response. Likewise, a deletion can be missed if the assay detects only sequence variants and does not adequately assess copy number. The genotype-to-score calculation can be arithmetically correct while resting on incomplete genetic data.

Apply CYP2C19-Specific Diplotype Rules

CYP2C19 illustrates why one should not force every gene into the same numerical scoring formula. For the common alleles listed here, use the specified gene-specific assignments. These are phenotype lookup rules; they do not define a universal additive scale on which *17 must be assigned a particular number. For less common alleles, follow a curated gene-specific reference rather than inferring a category from the allele name alone.

CYP2C19 copy-number changes require additional care. Unlike the commonly used CYP2D6 activity-score procedure, there is not a universal, routine CYP2C19 additive score that permits every duplication to be converted directly into a phenotype category. A confirmed deletion can indicate loss of a gene copy, but interpretation depends on the allele structure, assay result, and applicable guidance. Do not infer that duplication of a CYP2C19 allele automatically changes the phenotype category unless a validated gene-specific rule supports that assignment.

From Phenotype to PBPK Parameter

The phenotype label can help define a population stratum or identify subjects expected to differ in enzyme function, but a PBPK model needs an explicit parameterization. A modeler may use the genotype-derived result to modify the relevant enzyme capacity or intrinsic clearance, provided the mapping is supported for the drug and model. The size of that modification should come from appropriate functional or clinical evidence, not from treating a phenotype label as a numeric fold-change.

This distinction matters because whole-body drug disposition also depends on factors such as hepatic delivery, binding, and other clearance pathways. A genotype-linked change in one enzyme does not by itself specify the change in total clearance or exposure. The earlier material on genetic variation as a PBPK input motivates this separation: the genotype is an input to a mechanistic representation, not a substitute for the representation.

Diagnostics, Alternatives, and Common Pitfalls

Before using a translated activity value, inspect the path from genetic call to model input. Check that the allele names use the same reference as the activity values; confirm that a reported copy-number result identifies which allele is duplicated; verify the arithmetic and threshold boundary; and record the guideline or scoring version. If a score is close to a phenotype cutoff, report that sensitivity rather than implying a categorical assignment is biologically exact. If the allele is uncharacterized or the gene structure is unresolved, retain the uncertainty instead of forcing a confident label.

There are two broad ways to represent the result. A quantitative activity score, as commonly used for CYP2D6, preserves distinctions between diplotypes and can be useful when a model requires a graded functional input. A categorical phenotype lookup, commonly used for CYP2C19, applies curated gene-specific rules and is straightforward to communicate, but it may discard within-category variation. Neither approach is universally superior: use the method validated for the gene and purpose, and make any subsequent model parameter mapping explicit.

• Mistake: “There are two alleles, so there are always two copies.” Correction: deletions and duplications alter copy number; resolve them when relevant.
• Mistake: “A duplication always increases activity.” Correction: the effect depends on which allele is duplicated. Extra no-function copies do not contribute normal activity under the scoring convention.
• Mistake: “CYP2D6 and CYP2C19 use the same activity-score arithmetic.” Correction: CYP2D6 commonly uses an additive score with specified thresholds; CYP2C19 is often assigned by gene-specific diplotype rules.
• Mistake: “Normal metabolizer means normal clearance of every substrate.” Correction: the label summarizes inherited enzyme function under a defined system. Substrate dependence and other physiological or drug-related influences remain relevant.
• Mistake: “The phenotype label is a PBPK parameter.” Correction: state and justify the separate mapping from genotype-derived function to enzyme abundance, intrinsic clearance, or another model parameter.