# A gene-expression result is not a list of DNA changes

Draft • Research checked 4 October 2026 • AI-assisted DomDNA editorial content. No independent clinical review or publication is claimed. General education, not individual medical advice.

A headline says that a gene is “switched on”. A DNA report lists a variant in that gene. The two statements can be connected in research, but they are not interchangeable measurements. Gene expression concerns the use of genetic information, while a sequence result concerns the DNA letters examined.

This matters when a claim jumps from “associated with expression” to “your gene is active”. A person's DNA file does not, by itself, measure current RNA or protein levels in a particular tissue.

## Name what was measured

The [NHGRI gene-expression glossary](https://www.genome.gov/genetics-glossary/Gene-Expression) describes the production of RNA and other functional outputs from genetic information. Expression is regulated and can vary across cell types and conditions.

The familiar switch metaphor is useful only up to a point. A measured difference can concern how much RNA is present, which transcript is used or another defined output. It need not mean that the entire gene is wholly on or wholly off.

When reading a study, write down the measured material. Was the analysis of DNA sequence, RNA abundance, a protein or a functional activity? Those are different observations, even if the authors discuss the same gene.

The [NHGRI transcriptome fact sheet](https://www.genome.gov/about-genomics/fact-sheets/Transcriptome-Fact-Sheet) explains that the transcriptome concerns RNA products. It provides conceptual context rather than evidence that any consumer routine can control a chosen expression pattern.

## A hypothetical comparison across tissues

Imagine an editorial example in which researchers measure an RNA product in two tissue types. The amount differs, even though the cells share much of the same inherited DNA. The difference is not evidence that one tissue has lost the gene.

Now imagine a separate study linking a sequence variant with average expression differences in a tissue dataset. That relationship can help researchers investigate regulation. It does not directly measure the RNA amount in a new person carrying the variant.

The tissue label cannot be dropped without changing the claim. A result about one sampled tissue may not apply to another, and a measurement under one set of conditions may not describe every point in time.

Avoid treating the examples as an instruction to order RNA tests. They illustrate why the measurement and context should remain visible when a study is summarised.

## What an expression atlas can show

The primary [GTEx atlas paper](https://pmc.ncbi.nlm.nih.gov/articles/PMC7737656/) examines genetic regulatory associations across human tissues using donor samples. Its abstract describes RNA sequencing from post-mortem donors and analyses of expression and splicing.

That is a substantial research resource. It is not a live expression readout for a consumer whose DNA has been uploaded elsewhere. Nor is it a trial of an intervention intended to improve symptoms.

A useful interpretation of an atlas entry would identify the tissue, endpoint and association. An overextended interpretation would say that a person should change a supplement because a variant proves their gene is currently underactive.

We have checked the accessible paper's abstract and relevant explanatory text, not reproduced every analysis or independently assessed all donor and laboratory details. No specific variant or gene is selected for personal advice here.

## Separate regulation from benefit

Research showing a biological change can be important without showing a health benefit. An intervention might alter a measured expression pattern while leaving the outcome people care about untested.

If a product uses “gene activation” as its central promise, ask which output was measured, in what material and under what conditions. Then ask whether the study measured the claimed clinical or everyday benefit. A molecular endpoint should not stand in for an unmeasured outcome.

The reverse is also useful: an effective action does not need an elaborate gene-expression story to justify a practical benefit. The relevant outcome evidence should carry the claim.

## What you can do with this

Annotate one gene-expression headline with four fields: measured output, tissue or cell type, study setting and outcome. Keep a sequence association in a separate field from an expression measurement.

Your next question can be, “Was expression measured in these participants, or predicted from their DNA using another dataset?” That distinction often clears up a confusing report.

Understanding expression means knowing what the word refers to, not treating every gene-related statement as a personal instruction.

For general lifestyle learning, [DomDNA’s educational quiz](https://domdna.com/quiz) does not measure gene expression or infer it from your answers.