What HLA-B*57:01 Means in Your 23andMe Raw Data

Your 23andMe raw data contains SNP calls, not HLA allele names. When a third-party tool reports HLA-B57:01, some tools infer it from the `rs2395029` tag SNP in the HCP5 gene. Some tools may use multiple HLA markers or imputation. The G allele at `rs2395029` is a tag for HLA-B57:01. That association is best studied in European-descent populations, but it is a clue, not a clinical result.

What does HLA-B*57:01 mean in my 23andMe raw data?

HLA-B57:01 is one version of the HLA-B immune protein. Your 23andMe file does not have a row labelled HLA-B57:01. What you can see is rs2395029.

• Open the raw data file and search for rs2395029.
• G allele absent at rs2395029 means no copies of the tag allele were detected. Some tools report this as not flagged.
• G allele present at rs2395029 means at least one copy of the tag allele was detected. Any G allele copy is enough for some tools to flag it.
• A missing row can mean the SNP was not assayed on that 23andMe chip version or was not called. It must not be read as negative.

If a report says HLA-B*57:01 positive from 23andMe data, it may be saying rs2395029 G allele positive.

A G allele at rs2395029 is a clue that HLA-B57:01 may be present. It is not the same as a clinical HLA-B57:01 test. A consumer raw-data flag is not a prescription decision by itself. Abacavir decisions require clinician-supervised HLA-B*57:01 testing with a validated assay.

One G allele copy is enough

For abacavir risk, one HLA-B*57:01 allele is enough. At the rs2395029 marker, one G allele copy and two G allele copies are both treated as a possible-carrier flag. Two copies do not mean the reaction would be twice as bad. It just means both copies of the marker match the tag.

Why abacavir turns dangerous in HLA-B*57:01

HLA-B*57:01 is a protein that displays small pieces of other proteins to immune cells. Abacavir fits into the display groove and changes the set of peptides shown. Some of the new peptide-HLA combinations look foreign to T cells, which then attack. That attack produces a hypersensitivity reaction.

The reaction often looks like flu: fever, rash, tiredness, nausea, vomiting, abdominal pain, cough and sore throat. It usually appears in the first six weeks of treatment, often in the first ten days. Continuing the drug makes it worse. If abacavir hypersensitivity is suspected, stop the drug and do not restart it except under specialist supervision after the diagnosis is clarified. Rechallenge can be dangerous.

If you are not taking abacavir, does it matter?

Day to day, no. HLA-B57:01 is not a disease gene. It does not make you unwell on its own. It matters mainly if you are ever prescribed abacavir. The same HLA type has also been associated with liver injury from flucloxacillin, an antibiotic, so it is worth mentioning if that drug is ever considered. Most people with HLA-B57:01 never have an abacavir reaction because they never take the drug.

When to get a confirmatory HLA-B*57:01 test

If you have HIV and abacavir is being considered, clinical guidelines already call for HLA-B*57:01 testing with a validated assay before starting. A 23andMe result does not replace that.

• G allele present at rs2395029: ask for confirmatory HLA typing if abacavir is possible.
• G allele absent at rs2395029: do not treat it as clearance in a medical setting. A negative rs2395029 result is less reassuring outside European-descent populations because tag-allele linkage can differ.
• A missing call: treat it as not assayed or not called, not negative.
• Already had a reaction to abacavir: treat it as an allergic reaction and get proper care, whatever the raw file says.

The limits of a tag SNP

rs2395029 is a tag, not the HLA allele itself. It travels with HLA-B*57:01 often enough to be useful, but the strength of that link is best described in European-descent populations. In other ancestries the correlation can be weaker, so a negative rs2395029 result is less reassuring in those populations. Consumer chips also make genotype errors and no-calls. A tag can be right in a population and wrong in an individual.

Clinical HLA typing reads the actual HLA-B gene, not just a nearby marker. If the result will change a drug decision, confirm it.

healthosx.com/dna

The download steps for 23andMe, on the page itself.

healthosx.com/dna

The short version: what is worth acting on, first.

The free DNA tool at https://healthosx.com/dna reads a raw 23andMe, AncestryDNA, MyHeritage or whole-genome VCF file in the browser. The immediate no-account browser view shows the initial trait and health story, not necessarily the medication-response report. The medication-response report is part of the free full report that opens after a one-time email code. It includes an HLA-B medication-response finding, such as whether one HIV medicine is safe according to the raw-data flag, but that is not clinical clearance or a prescription decision.