Can the 3' rule cause a change in variant reference allele sequence?

7 views
Skip to first unread message

Manuel Luypaert

unread,
Apr 27, 2026, 2:17:25 PM (3 days ago) Apr 27
to HGVS Nomenclature
Hi all,

I have a question about how to interpret "reference sequence" in the 3' rule and what the possible effect can be on the variant's reference allele sequence. As an example:

Given the (let's say genome) reference sequence  TAGCTAGCTAGCTA, and the variant which deletes AGCT at positions 6-9 (g.6_9del), what is the correct HGVS representation for this variant?  g.10_13del, keeping the reference allele sequence as AGCT? Or g.11_14del, changing the variant's reference allele sequence to GCTA? Both result in the same end product on inclusion in the reference genome sequence provided above.

Best regards,
Manuel

Reece Hart

unread,
Apr 29, 2026, 5:05:13 PM (yesterday) Apr 29
to Manuel Luypaert, HGVS Nomenclature
Hi Manuel-

Yes, a variant that inserts or deletes sequence in the context of a repeated region is subject to 3' normalization (also called shifting or shuffling). 

I recommend avoiding the phrase "reference allele sequence" in this context to avoid confusion with the subsequence corresonding to the variant itself. (In HGVS Nomenclature terminology, "reference sequence" is the proper name for the entire sequence. It is, of course, the sequence for a specific biological allele.)

And, perhaps to state the obvious, the proposition of your statement "the variant which deletes AGCT at positions 6-9" is typically not actually known with that precision: 4 nucleotides in that repeat were deleted, but a deletion of any sequence of 4 of them would have the same result. Folks at NCBI called this mirage "overprecision", and it's why some practitioners prefer "fully justified normalization", which creates a bubble variant that covers the entire region of position uncertainty.

Some of these links might help:

-Reece


--
You received this message because you are subscribed to the Google Groups "HGVS Nomenclature" group.
To unsubscribe from this group and stop receiving emails from it, send an email to hgvs-nomenclat...@googlegroups.com.
To view this discussion visit https://groups.google.com/d/msgid/hgvs-nomenclature/89790869-1827-478c-8d34-b3346deefd30n%40googlegroups.com.
Reply all
Reply to author
Forward
0 new messages