Why the Nearest Gene Is Not Necessarily the Target

Introduction

When a genetic locus is close to a gene, that gene is an obvious candidate. But “closest” describes a position on a chromosome; it does not, by itself, identify the gene through which a variant acts. A regulatory element can contact and influence a more distant promoter, and a locus can have different plausible targets in different tissues. Gene orientation and promoter choice add further reasons that a simple nearest-gene label can mislead.

The aim here is not to replace one automatic assignment rule with another. It is to diagnose when proximity is a weak or incomplete basis for attribution, and to state what regulatory evidence does—and does not—support. A link from a variant-containing regulatory element to a gene is evidence for a candidate relationship, not proof that the variant changes that gene or causes a trait through it.

What “Nearest Gene” Measures

An annotation may instead define “nearest gene” by distance to the gene body, such as the shortest distance to any base in the gene’s genomic span. Those conventions can yield different answers, especially when a variant falls within a gene or when neighboring genes overlap. A report should therefore say which distance was used. In the examples below, nearest-gene distances are measured to the nearest annotated TSS.

Coordinates must refer to the same genome build, and the candidate genes and regulatory elements must be annotated on that build. A distance calculated across mismatched coordinates is not meaningful. Even with correct coordinates, distance is only a description of genomic layout. It does not tell us whether the variant overlaps a regulatory element, whether that element is active in a relevant tissue, or which promoter it may regulate.

Orientation Is Not a Target Assignment

A gene’s strand determines the direction in which its sequence is transcribed. On the plus strand, transcription proceeds toward increasing genomic coordinates; on the minus strand, it proceeds toward decreasing coordinates. A promoter is generally near a gene’s TSS, but “upstream” depends on the gene’s strand: it is toward lower coordinates for a plus-strand gene and toward higher coordinates for a minus-strand gene.

This matters when interpreting a locus between genes. A variant can be upstream of one gene and downstream of another, but those labels do not establish which gene is regulated. Regulatory elements can act over a distance and need not lie on the promoter’s upstream side. Nor does a promoter’s orientation reveal which enhancer contacts it. Orientation helps interpret the map; it is not a substitute for functional evidence.

Distal Links and Promoter Choice

A regulatory-element-to-gene link changes the question from “Which gene is closest?” to “Which promoter, if any, is supported as a potential target of the element?” Earlier material on promoters, enhancers, chromatin accessibility, and chromatin contact provides the relevant evidence types. An enhancer-like chromatin state can support the interpretation that a region is regulatory; a contact can support a physical relationship between that region and a promoter. The value of those observations depends partly on whether they come from the tissue or cell type relevant to the locus.

A gene can also have multiple promoters and TSSs. If the variant-containing element contacts one promoter, it is not enough to say only that it “links to the gene” when the promoter distinction matters. Different promoters can be active in different tissues or produce different transcript forms. A nearest-TSS calculation may select one promoter because it is closest, while regulatory evidence points to another promoter of the same gene—or to a promoter of a different gene.

Why Tissue Matching Matters

Regulatory activity and chromatin contacts can vary by tissue and cell type. A contact observed in a cell type relevant to the phenotype is generally more pertinent to interpreting the locus than a contact observed only in an unrelated cell type. This is a relevance judgment, not a claim that a contact in a relevant tissue proves a mechanism. Likewise, failure to observe a contact can reflect assay coverage or sensitivity, as well as a genuinely absent interaction.

When comparing evidence, keep its source and context visible. An enhancer mark in one tissue and a contact in another do not jointly demonstrate that the element is active and connected to the promoter in the same biological context. If evidence comes from different tissues, describe that mismatch rather than combining the observations as though they were a single coherent chain. The earlier material on tissue and cell-type matching and missingness is useful here: “not observed” and “observed to be absent” are not interchangeable.

A Short Justification: Distance Alone Cannot Identify a Target

The limitation of nearest-gene attribution follows directly from how the rule is defined. Let \(G\) be the set of annotated genes and \(t_g\) the TSS coordinate chosen for gene \(g\). The rule returns:

The inputs to this calculation are the variant coordinate and the TSS coordinates. There is no term for enhancer activity, contact, tissue, or promoter usage. Consequently, two loci with the same variant and TSS coordinates receive the same nearest-gene label even if regulatory evidence differs between them. For example, adding a tissue-matched contact from a variant-containing enhancer to a more distant promoter does not change any \(|v-t_g|\), and therefore cannot change \(g_{\mathrm{near}}\). It does change the functional evidence available for interpreting the locus. Thus, the nearest-gene rule can be a useful positional summary, but it cannot, on its own, establish a regulatory target.

A Practical Diagnostic

When a nearby gene is being treated as the target, check whether the conclusion depends only on its distance. A disciplined review separates positional description from functional support:

Common Mistakes and Why the Distinction Matters

Nearest-gene attribution is convenient because it is reproducible and easy to calculate. Its convenience can make it seem stronger than it is. Treating the nearest gene as the target without checking regulatory evidence can misdirect follow-up experiments, obscure a distal candidate, or conceal a promoter-specific interpretation. The correction is not to assume that the farthest contacted gene must be correct: contacts and regulatory annotations also have limitations. Instead, describe the positional result and the functional evidence separately.

• Wrong reasoning: “The variant is closest to Gene A, so Gene A is the target.” Correction: It is nearest to Gene A under a stated distance rule; target status requires other relevant evidence.
• Wrong reasoning: “The variant is upstream of Gene B, so it regulates Gene B.” Correction: Upstream depends on strand orientation and does not establish a regulatory link.
• Wrong reasoning: “The element contacts Gene C, so the variant definitely changes Gene C.” Correction: A contact supports a candidate relationship, but does not prove that the variant alters regulation or causes a trait through that gene.
• Wrong reasoning: “There is no contact in the available data, so the gene cannot be targeted.” Correction: A missing observation may reflect incomplete coverage or assay limitations.
• Wrong reasoning: “A contact in any cell type is equally informative.” Correction: Report the tissue and judge its relevance to the biological question.