23andMe Raw Data Plus Strand Orientation Explained
- Aug 1
- 5 min read
23andMe raw data uses the plus strand of the specified reference genome assembly. By default, 23andMe displays genotypes relative to the positive strand of GRCh37, also called Build 37 [R]. Its Browse Raw Data tool can also display results relative to the positive strand of GRCh38 [R].
This matters because research papers, clinical databases, and DNA interpretation tools may describe the same variant using the opposite DNA strand, a gene transcript, or a different genome assembly. As a result, two sources can show different allele letters without actually disagreeing about the underlying genetic result.
The direct answer: 23andMe uses the plus strand
Each chromosome contains two complementary strands of DNA:
a positive, or plus, strand
a negative, or minus, strand
23andMe states that it reports genotypes on the positive strand of the reference assembly being displayed. Its default raw-data view uses GRCh37, while GRCh38 is also available in the Browse Raw Data interface.
This means that a genotype shown by 23andMe should not be compared directly with allele letters from another source until that source’s strand orientation and genome assembly are known.
Already have your 23andMe raw DNA file?
GenesUnveiled checks supported markers using a consistent allele orientation and organizes the findings available in your file.
Client-side analysis · Free account required
What plus and minus strands mean
The two strands of DNA are complementary. Each DNA base pairs with one specific base on the opposite strand:
Plus-strand base | Complementary minus-strand base |
A | T |
T | A |
C | G |
G | C |
When a source reports the opposite strand, the genotype letters must be complemented.
For example:
plus-strand GG corresponds to minus-strand CC
plus-strand AA corresponds to minus-strand TT
plus-strand AG corresponds to minus-strand CT
plus-strand AC corresponds to minus-strand GT
The order of two genotype letters is not significant. AG and GA represent the same heterozygous genotype.
Why your 23andMe result may not match another website
Several forms of notation can make the same variant appear different.
Opposite strand orientation
A publication may describe alleles from the negative strand while 23andMe displays the positive strand.
For example, 23andMe’s official documentation explains that a GG result on the positive strand corresponds to CC when another source reports the same marker from the negative strand.
Neither result is necessarily incorrect. They may simply describe complementary DNA strands.
Genomic versus transcript notation
Genes can be located on either genomic strand. Clinical sources often describe a variant relative to the gene’s messenger-RNA transcript rather than the chromosome’s positive strand.
This can produce different-looking allele changes even when the rsID is identical.
Different reference genome assemblies
A genomic position is meaningful only in relation to a reference assembly such as:
GRCh37
GRCh38
The rsID may remain the same while the chromosome position changes between assemblies. In a small number of genomic regions, the orientation itself may also differ between assemblies.
For this reason, a reliable comparison should check:
the rsID;
the genome assembly;
the strand orientation;
the allele letters;
the type of notation being used.
Historical variant names
Some well-known variants still appear under older names based on previous transcript numbering or naming conventions. These names may coexist with current genomic, transcript, protein, and rsID descriptions.
Factor V Leiden: a clear strand-orientation example
Factor V Leiden rs6025 illustrates why allele letters must be interpreted in context.
ClinVar describes the variant as [R]:
genomic GRCh38 notation: C>T
F5 transcript notation: c.1601G>A
protein notation: p.Arg534Gln
historical protein name: R506Q
rsID: rs6025'
The genomic C>T and transcript G>A descriptions refer to the same variant. The letters differ because the F5 transcript is read from the complementary genomic strand.
Someone comparing a C/T result in a consumer raw DNA file with a clinical source describing G/A could incorrectly conclude that the sources conflict. Correct strand handling shows that the alleles are complementary.
You can read more in the GenesUnveiled guide to Factor V Leiden rs6025 and blood-clot risk.
Is the plus strand always the same as the forward strand?
The terms plus strand and forward strand are often treated as equivalent in consumer genetics, but technical databases have not always used “forward” in exactly the same way.
The plus strand refers to the positive strand of a specified genome assembly. Historically, dbSNP could use “forward” to mean the orientation of an allele relative to its submitted reference sequence, which was not necessarily identical to the genomic plus strand.
Modern NCBI RefSNP reporting presents alleles forward relative to the specific sequence placement being shown, helping reduce older orientation ambiguity [R]. For practical raw-DNA interpretation, “this is the forward allele” is not enough by itself. The source should also identify the sequence or genome assembly to which that orientation refers.
Why A/T and C/G SNPs need extra care
Some SNPs have the allele pairs:
A/T
C/G
These are sometimes called strand-ambiguous or palindromic SNPs. Complementing A/T still produces A/T, while complementing C/G still produces C/G.
The allele pair alone therefore cannot reveal whether one source has flipped the strand. Additional information—such as surrounding sequence, genome position, reference assembly, population frequency, or explicit strand annotation—may be needed.
This is one reason automated DNA analysis should not infer strand orientation solely from the two allele letters.
What does “--” mean in a 23andMe raw file?
A downloaded 23andMe file may display -- instead of a two-letter genotype.
This means that the marker was not determined, also called a no-call. It does not mean:
that the person has the common genotype;
that the variant is absent;
that both alleles are missing from the genome;
or that the result should be interpreted as normal.
23andMe uses a not-determined result when its algorithm cannot make a sufficiently confident genotype call at that marker. Some no-calls are expected in a consumer raw DNA file.
A responsible analysis platform should treat a no-call as unavailable data rather than assign a genotype.
How GenesUnveiled handles strand orientation
GenesUnveiled uses a defined forward genomic orientation for supported markers.
When a research paper or database reports alleles from the opposite strand, the allele direction must be converted before the finding can be mapped to a raw DNA result. This includes complementary conversions such as:
source A → forward T
source T → forward A
source C → forward G
source G → forward C
The platform also distinguishes between:
a supported genotype;
a marker missing from the file;
a no-call;
and a result that cannot be interpreted reliably.
This reduces the risk of assigning the effect of one allele to its complementary counterpart. GenesUnveiled’s research and review process is described in How We Build DNA Reports. During analysis, your raw file and per-variant results remain client-side in the user’s browser.
Comments