Nomenclature recurrent CNV

[Rikke Christensen]

Dear HGVS,

 

We have discussed how we correctly write the nomenclature for a recurrent CNV identified by whole genome sequencing.

 

We have e.g. identified the classic 22q11 (LCR22A-LCR22D) deletion in a patient by WGS.

 

In the HGVS homepage there is the following example:

 

• NC_000023.11:g.(86200001_103700000)del
  for ISCN del(X)(q21.31q22.2)
  (within a chromosome, breakpoint not sequenced)

 

We are in doubt if we can interpret this as “breakpoint not properly mapped” instead of “breakpoint not sequenced” , because when we perform WGS the breakpoint is sequenced, however due to the LCR regions flanking the recurrent CNV the data are not properly mapped in those regions in the hg38 reference. It is often not possible to identify the exact breakpoint. We therefor think it is a good idea to use the brackets to indicate in the nomenclature that we do not know the exact breakpoints. Would it be correct to write like this:

 

NC_000022.11:g.(18747401_21224600)del  (here we use the breakpoints given by the bioinformatic algorithm that are approximate as we cannot see the breakpoint exactly int the software).

 

Or would it be better to give an interval for the two breakpoint regions and then write like this:

 

NC_000022.11:g.(18159723_18924717)_(21111384_21562827)del

 

We look forward to hear your opinion on this.

 

Kind regards, Rikke

 

 

Rikke Christensen

Clinical Seniorspecialist, PhD

Clinical Laboratory Geneticist (CLG)

Tel. +45 7845 5506

rikk...@rm.dk

Department of Clinical Genetics

Brendstrupgaardsvej 21C ▪ DK-8200 Aarhus N

 

[Johan den Dunnen]

Dear Rikke,

the description you refer to, "NC_000023.11:g.(86200001_103700000)del" is unfortunately outdated and needs to be corrected. For deletions where the breakpoint has not been sequenced the HGVS nomenclature uses a 4-postion format like g.(123456_234567)_(345678_456789)del, indicating the uncertainty about the location of the deletion breakpoints, where (123456_234567) and (345678_456789)  indicate the two regions where the break points should be located (see https://hgvs-nomenclature.org/stable/recommendations/uncertain/). For the example given this gives the description NC_000023.11:g.(87000001_92700000)_(103300001_104500000)del.

Getting to your case, yes the recommendation is to give an interval for the two breakpoint regions and write the variant as NC_000022.11:g.(18159723_18924717)_(21111384_21562827)del.

Best regards,

Johan den Dunnen
HUGO HGVS Variant Nomenclature Committee (HVNC)

Op dinsdag 4 november 2025 om 15:09:14 UTC+1 schreef Rikke Christensen:

[Laura Conlin]

Hi Rikke,

If you would like to use ISCN nomenclature to describe a copy number finding where you do not have exact breakpoints, this is how you would format your finding:

seq[GRCh38] 22q11.21(18159723x2,18924717_21111384x1,21562827x2)

This nomenclature describes the copy number state of the last "normal" base that you are sure is 2 copies, then the region that you are sure is one copy, followed by the next base where "normal" 2 copies begins again. The breakpoints would then fall in-between the first two bases (start), and between the last two bases (end). This format indicates that only copy number information is known by using the x before the copy number state (with x1  instead of a "del"). This is the format that ISCN uses for sequencing data where you ONLY have copy number state and have not been able to determine the precise breakpoint NOR have been able to determine if there was any rearrangements associated with the copy number change from information like aberrant mate pairs. If this was a copy number gain instead of a loss and you did NOT know if the additional copy was located in a tandem orientation (as "dup" in HGVS means tandem) or inserted elsewhere in the genome, then you can use a similar ISCN format, just with x3 instead of x1.

What is listed in the example you copied from the HGVS website is not quite correct for ISCN. The ISCN committee had proposed a dual nomenclature that if you would like to describe the sequence finding at the chromosome level in addition to the breakpoint information, then you can use a combined nomenclature where the first line is the ISCN using the "karyotype" format to indicate the overall structural finding with the HGVS portion describing the breakpoint in the next line. The ISCN portion is not meant to be stand alone, but used as a supplement describing the gross chromosomal information in a way consistent with karyotypes. The HGVS portion can be stand alone if you do not want to describe the overall structure; however, "del" in HGVS means (at least to me) that you DO know that this is JUST a deletion, and it is NOT a part of a rearrangement. I find this combined format most helpful when there IS a rearrangement (like a translocation or a recombinant chromosome), and I stick to the ISCN-only nomenclature (as above) to describe copy number variants that do not have structural information (esp for recurrent CNVs where we did not capture breakpoints/structural information).

Combined ISCN/HGVS to describe a non-complex, straightforward deletion in an XX individual (with the normal X listed first to indicate if this deletion is hemizygous or heterozygous)

seq[GRCh38] X,del(X)(q21.31q22.2)

NC_000023.11:g.(86200001_103700000)del

Laura (ISCN Committee member)

[Johan den Dunnen]

Dear Laura,

please note that the HGVS part of the combined ISCN/HGVS description you give is not correct. Correct is:

seq[GRCh38] X,del(X)(q21.31q22.2)

NC_000023.11:g.(87000001_92700000)_(103300001_104500000)del.

Best regards,

Johan den Dunnen
HUGO HGVS Variant Nomenclature Committee (HVNC)

Op woensdag 5 november 2025 om 23:30:04 UTC+1 schreef lkco...@gmail.com:

[Laura Conlin]

HI Johan,

My example was meant to show a combined ISCN/HGVS for a straight-forward deletion, meaning there was no ambiguity at the breakpoint (you can captured the breakpoint exactly). Sorry if this was not clear.

Laura

[Johan den Dunnen]

Dear Laura,

when the deletion breakpoint has been sequenced, the format is NC_000023.11:g.86200001_103700000del, i.e. without "( )". In HGVS nomenclature parenthesis indicates "uncertainty". When the breakpoint has been sequenced, there is no uncertainty.

Best regards,

Johan den Dunnen
HUGO HGVS Variant Nomenclature Committee (HVNC)

Op woensdag 19 november 2025 om 18:19:39 UTC+1 schreef lkco...@gmail.com: