Describing variants

"mutation nomenclature"

recommendations for the

description of DNA changes

Johan den Dunnen

chair SVD-WG

VarNomen @ HGVS.org
http://www.HGVS.org/varnomen/

© JT den Dunnen
 HGVS / HVP / HUGO
Sequence Variant Description
working group
Working Group Members:
• Anne-Francoise Roux (EGT)
• Donna Maglott (NCBI/EBI)
• Jean McGowan-Jordan (ISCN)
• Peter Taschner (LSDBs)
• Raymond Dalgleish (LSDBs)
• Reece Hart (industry)
• Johan den Dunnen (chair)
• HGVS - Marc Greenblatt
• HUGO - Stylianos Antonarakis

© JT den Dunnen
 Nomenclature
( describing DNA variants )

Stable

Meaningful Memorable

Unequivocal

© JT den Dunnen
 Definitions
• prevent confusion
do not use "mutation"
use variant, disease-associated variant
do not use "polymorphism"
use variant, not disease-associated variant
do not use "pathogenic"
use disease-associated, a disease-associated variant

• better use neutral terms

sequence variant
alteration
CNV (Copy Number Variant)
SNV (Single Nucleotide Variant, not SNP)

© JT den Dunnen
 Variant description
the basis http:// www.HGVS.org / varnomen

Hum Mutat (2016) 37:564-569

© JT den Dunnen
www.HGVS.org/varnomen
Follow the recommendations

when you disagree, start a debate -do not use
private rules, this only causes confusion

© JT den Dunnen
facebook & twitter

© JT den Dunnen
 Versioning

version presented is 15.11 (Nov.2015)

© JT den Dunnen
 Variant types
• change in sequence
ACATCAGGAGAAGATGTTC GAGACTTTGCCA
ACATCAGGAGAAGATGTTT GAGACTTTGCCA
ACATCAGGAGAAGATGTT GAGACTTTGCCA
ACATCAGGAGAAGATGTTCCGAGACTTTGCCA

• change in amount

Structural Variation (SV)

(Copy Number Variation)

• change in position

© JT den Dunnen
DNA, RNA, protein
• unique descriptions
prevent confusion

• DNA
A, G, C, T
g.957A>T, c.63-3T>C

• RNA
a, g, c, u
r.957a>u, r.(?), r.spl?

• protein ( mostly deduced )

three / one letter amino acid code
* = stop codon
p.(His78Gln)

© JT den Dunnen
Reference sequence
• use official HGNC gene symbols
• provide reference sequence
covering complete sequence
largest transcript
preferably a LRG
e.g. LRG_123
give accession.version number
e.g. NM_012654.3

• indicate
DNA
type of Reference Sequence
coding DNA c. genomic g.
mitochondrial m. non-coding RNA n.
RNA r.
protein p.

© JT den Dunnen
 The LRG

EBI, NCBI, Gen2Phen

© JT den Dunnen
Numbering residues
• start with 1
genomic 1 is first nucleotide of file
no +, - or other signs
coding DNA 1 is A of ATG
for introns refer to genomic Reference Sequence

• repeated segments
assume most 3' as changed
( ...CGTGTG TG A… )

• coding DNA only

5' of ATG …, -3, -2, -1, A, T, G, …
no nucleotide 0
3' of stop *1, *2, *3, …
no nucleotide 0
intron
position between nt's 654 and 655
c.654+1, +2, +3, ………, -3, -2, c.655-1
change + to - in middle

© JT den Dunnen
 Numbering
• RNA ( deduced mostly )

like coding DNA

• protein ( deduced only )

from first to last amino acid

rule of thumb: c. nucleotide position divided
by 3 roughly gives amino acid residue
description between parantheses

© JT den Dunnen
 Reference Sequence
coding DNA reference sequence (c.)

non-coding DNA reference sequence (n.)

© JT den Dunnen
coding DNA or genomic ?
• human genome sequence
complete
covers all transcripts
different promoters, splice variants, diff.
polyA-addition, etc.
but
hg19 chr2:g.121895321_121895325del
is long & complicated
huge reference sequence files
new builds follow each other regularly
carries no understandable information

• coding DNA
does not cover all variants
but gives a clue towards position

© JT den Dunnen
Numbering - genomic 3

• g.12158663A>G
• g.23669859>C
• g.89112396G>A no relation to

• g.112775623C>G RNA & protein

• g.56569443A>T
• g.12741333T>G
• g.188153979G>C
© JT den Dunnen
Numbering - coding DNA
• c.1637A>G
protein coding region

• c.859+12T>C
in intron (5' half) relation to

• c.2396-6G>A
in intron (3' half)
RNA & protein

• c.-23C>G
5' of protein coding region (5' of ATG)

• c.*143A>T
3' of protein coding region (3' of stop)

• c.-89-12T>G
intron in 5' UTR (5' of ATG)

• c.-649+intron
79G>C
in 3' UTR (3' of stop)
© JT den Dunnen
Types of variation
• simple
substitution c.123A>G
deletion c.123delA
duplication c.123dupA
insertion c.123_124insC
other
conversion, inversion, translocation, transposition

• complex
indel c.123delinsGTAT

• combination
two alleles
of variants
c.[123A>G];[456C>T]
>1 per allele c.[123A>G;456C>T]

© JT den Dunnen
 Substitution
• substitution designated by ">"
> not used on protein level

• examples
genomic g.54786A>T
cDNA c.545A>T
( NM_012654.3 : c.546A>T )

RNA r.545a>u

protein p.(Gln182Leu)

© JT den Dunnen
 Deletion
• deletion
designated by "del"
range indicated by "_"

• examples
c.546del
c.546delT

c.586_591del
c.586_591delTGGTCA, NOT c.586_591del6

c.(780+1_781-1)_(1392+1_1393-1)del
exon 3 to 6 deletion, breakpoint not sequenced

© JT den Dunnen
 Duplication
• duplication
designated by "dup"
range indicated by "_"

• examples
c.546dup
c.546dupT

c.586_591dup
c.586_591dupTGGTCA, NOT c.586_591dup6
do not describe as insertion

c.(780+1_781-1)_(1392+1_1393-1)dup
exon 3 to 6 duplication, breakpoint not sequenced
NOTE: dup should be in tandem

© JT den Dunnen
 Insertion
• insertion
designated by "ins"
range indicated by "_"
! give inserted sequence

• examples
c.546_547insT
NOT c.546insT or c.547insT

c.1086_1087insGCGTGA
NOT c.1086_1087ins6

c.1086_1087insAB567429.2:g.34_12567
when large insert submit to database and
give database accession.version number

© JT den Dunnen
 Inversion
• inversion
affecting at least 2 nucleotides
designated by "inv"
range indicated by "_"

• example
c.546_2031inv
NOT c.2031_546inv

© JT den Dunnen
 Conversion
• conversion
affecting at least 2 nucleotides
designated by "con"
range indicated by "_"

• examples
c.546_657con917_1028
c.546_2031conNM_023541.2:c.549_2034

© JT den Dunnen
Sequence repeats
• mono-nucleotide
g.8932A(18_23)
stretches
() = uncertain
c.345+28T(18_23)
alleles 345+28T[18];[21]

• di-nucleotide stretches
c.1849+363CAG(13_19)
c.1849+363_1849+365(13_19)

• larger
g.532_3886(20_45)
3.3 Kb repeat

© JT den Dunnen
 SNVs (SNPs)

• SNV's
at least once give description based
on genome reference sequence

hg19 chr9:g.3901666T>C

rs12345678:T>C
dbSNP entry

© JT den Dunnen
Characters & codes
• codes used
+, -, *
> substitution (nucleotide)
_ range
; separate changes (in/between alleles)
, more transcripts
() uncertain
[] allele
= equals reference sequence
? unknown
del deletion
dup duplication
ins insertion
inv inversion
con conversion
ext extension
fs frame shift

© JT den Dunnen
 Uncertainty breakpoints
• Copy Number Variants
( last-normal_first-changed ) _ ( last-changed_first-normal ) del

BAC / PAC probe

chrX:g.(32218983_32238146)_(32984039_33252615)del
hg19

SNP-array
chrX:g.(32218983_32238146)_(32984039_33252615)del
GRCh36.p2
(rs2342234_rs3929856)_(rs10507342_rs947283)del

© JT den Dunnen
Uncertainty breakpoints 2

• whole exon changes

c.(423+1_424-1)_(631+1_632-1)del
intragenic deletion

c.(?_-79)_(631+1_632-1)del
deletion incl. 5' end

c.(423+1_424-1)_(*763_?)del
deletion incl. 3' end

c.(?_-79)_(*763_?)del
whole gene deletion, start/end undefined

describe what was actually tested

© JT den Dunnen
 Alleles
• allele
indicated by "[ ]", separated by ";"

• 2 c.changes, 2 alleles
[428A>G] ; [83dupG]

• 1 c.allele, several changes

[12C>G ; 428A>G ; 983dupG]

• 2 c.changes, allele unknown

428A>G (;) 83dupG

• mosaicism
special cases spaces in
description used
c. 428A= / A>G for clarity only
chimerism
c. 428A= // A>G

© JT den Dunnen
 Complex
• deletion / insertions
"indel"

c.1166_1177delinsAGT

• descriptions may become complex

when only an expert understands the
"code" consider database submission
description: c.875_941delinsAC111747.1

© JT den Dunnen
Changes in RNA
• description like DNA
r. / a, g, c, u

• examples
r.283c>u
r.0 no RNA from allele
r.? effect unknown
r.spl affects RNA splicing
r.(spl?) may affect splicing
r.283= no change
(equals reference sequence)

r.[=, 436_456del]
two transcripts from 1 allele

© JT den Dunnen
Changes in RNA 2

• one allele, 2 transcripts

effect on splicing not 100%

c.456+3G>C

on RNA r.[=, 436_456del]

> p.[=, Arg146_Lys152del]

© JT den Dunnen
Changes in protein
• description like DNA
p. / Ala, Cys, Gly, His, …, Ter
p. / A, C, D, E, F, G, H, …, *

• examples
nonsense
no RNA data
r.(?)
p.Trp65* (p.W65* / p.Trp65Ter) p.(Trp56*)
no stop
p.*1054Glnext*31
p.0 - no protein
p.Met1? - likely, but unknown effect
NOT p.Met1Val
fs - frame shift

© JT den Dunnen
 Frame shifts
• short form (sufficient)
p.Arg83fs

• long from (more detail)

p.(Arg83Serfs*15) (no RNA analysis)

indicate
first amino acid changed do not try to include
position changes at DNA level
first changed amino acid
length shifted frame
(from first changed to * incl.)
do not describe del, dup, ins, etc.

© JT den Dunnen
Recent additions
• added versioning
to support users
easier to find latest changes
allows statement "following HGVS version 2.0"

• stricter definitions
separate different classes
added hierarchy
computer-generated description
automated error-checking ( Mutalyzer )

• simplified use special characters

"_", ";", "+", "*", …
improved consistency

© JT den Dunnen
Acknowledgement
Presentation prepared by:
Johan den Dunnen
Human Genetics & Clinical Genetics
Leiden University Medical Center
Leiden, Nederland

chair SVD-WG

date: April 2017

© JT den Dunnen