Transcript LRGs

Locus Reference Genomic
(LRG) Sequences
Raymond Dalgleish
Department of Genetics
University of Leicester
Background
• Descriptions of sequence variants should
use HGVS nomenclature
• Variants should be described with respect
to a reference DNA sequence specified by
an accession number and a version
e.g. NM_000088.3:c.2362G>T
• Mostly works well, but three key issues
frequently cause problems for LSDB
curators and for diagnostic laboratories
Issue 1: Version not specified
• The autosomal dominant RP10 form of
retinitis pigmentosa is caused by variants
in the IMPDH1 gene
• Variants for this gene are described with
respect to NM_000883.1, but the version
is rarely mentioned in the literature
• The current version (NM_000883.3)
records a shorter mRNA & protein which
could lead to confusion and delay
Issue 2: Alternative splicing
• ~93% of genes have alternatively spliced
transcripts & may yield several proteins
• The CDKN2A locus encodes the tumour
suppressor proteins p16INK4a and p14ARF
• The mRNAs for the two proteins share
exon 2 in common but in different reading
frames, due to different upstream exons
• Separate RefSeq records for the mRNAs
CDKN2A alternate splicing
Issue 3: Legacy numbering (1)
• The “sickle cell” variant of β-globin is due
to the substitution of glutamic acid by
valine at amino acid 6
• Determined by amino acid sequencing
prior to completion of the genetic code
• HGVS protein-level description is
p.Glu7Val counting from the start codon
Issue 3: Legacy numbering (2)
• Type I & III collagen variants were
originally numbered from the start of the
Gly-X-Y triple-helical repeat region
• Legacy and HGVS descriptions still run in
parallel: e.g. Gly610Cys & p.Gly788Cys
• The exons of these genes were originally
numbered in a 3´ to 5´ direction
Issue 3: Legacy numbering (3)
• New exons are often discovered in genes
long after their initial characterisation
• This interferes with simple sequential
numbering of exons from 5´ to 3´
• Non-simple numbering is well-established:
– COL1A1: 33/34
– CFTR: 6a, 6b,14a, 14b, 17a, 17b
– OPRM: O, X, Y
– CDKN2A: 1B, 1A
So what is the solution?
• An ideal reference sequence would:
– be stable over periods as long as 25 years
– be free of version confusion
– comprise an “idealised” genomic DNA
sequence haplotype providing a practical
working framework
– contain comprehensive information about the
transcripts and proteins encoded by the gene
(including alternative numbering schemes)
– be mapped to the current genome assembly
Primary design decisions
• LRGs will be a working representation of a
gene with a permanent ID: i.e. no versions
• Based on any existing RefSeqGene record
• 5 kb upstream and 2 kb downstream
• There can be more than one LRG for a
given region of the genome
• LRGs will have both fixed and updatable
feature annotations
Primary fixed annotations
• Coding sequence coordinates
• Transcripts essential to the reporting of
sequence variants
• The conceptual translated protein(s)
• Non-coding transcripts
Primary updatable annotations
•
•
•
•
Mapping to current genome assembly
Chromosome number
Any alternative IDs
Cross references to other reference
sequences
• “Legacy” exon and amino acid numbering
systems
• Links to LSDBs
• Overlapping genes
Variant reporting with LRGs
• The calcitonin gene (CALCA) encodes the peptide
hormones calcitonin and calcitonin gene related peptide
(CGRP) by alternative splicing
• A SNP in the first base of exon 4 affects the transcript
(t2) and the resulting precursor protein (p2) for calcitonin
• The variant can be reported at gene, mRNA and protein
level with reference just to LRG_13 (CALCA)
Description Level RefSeqGene or RefSeq
LRG
gene
NG_015960.1:g.8290C>A
LRG_13:g.8290C>A
mRNA
NM_001033952.2:c.228C>A
LRG_13t2:c.228C>A
protein
NP_001029124.1:p.Ser76Arg
LRG_13p2:p.Ser76Arg
Progress
• LRGs can be viewed at the LRG web site:
http://www.lrg-sequence.org
• The first 10 LRGs have been finalised:
– COL1A1, COL1A2, COL3A1, CRTAP, ATP1A2,
CACNA1A, SCN1A, PPIB, FKBP10, CALCA
• Another 4 await final approval:
– LEPRE1, CDKN2A, L1CAM, UBE3A
• Requests have been received for around
100 others
Other tools to view LRGs
• Ensembl, NCBI Genome Workbench,
NCBI Sequence Viewer will soon provide
support for LRGs
• NGRL Universal Browser displays LRGs
with links through to LSDBs and dbSNP
• Mutalyzer will be updated to parse LRGs
to support their use in LOVD
• Alamut will probably be the first
commercial software support for LRGs
How do I learn more?
• Dalgleish et al., 2010, Genome Medicine, in press
• LRG web site:
http://www.lrg-sequence.org
• LRG specification document:
http://www.lrg-sequence.org/docs/LRG.pdf
• The LRG XML schema is available for download
• E-mail addresses:
– Request help: [email protected]
– Provide feedback: [email protected]
– Request a new LRG: [email protected]
Acknowledgements
Raymond Dalgleish
Tony Brookes
University of Leicester, Leicester, UK
Donna Maglott
Alex Astashyn
Ray Tully
NCBI, Bethesda, USA
Fiona Cunningham
Paul Flicek
Ewan Birney
Yuan Chen
Pontus Larsson
Will McLaren
Glenn Proctor
Brendon Vaughan
EBI, Hinxton, UK
Andrew Devereau
Glen Dobson
Christophe Béroud
INSERM, Montpellier, France
NGRL, Manchester, UK
Peter Taschner
Johan den Dunnen
LUMC, Leiden, Netherlands
Heikki Lehväslaiho
CBRC, Jeddah, Saudi Arabia
Coordination and funding
• LRGs were devised by the GEN2PHEN
project: http://www.gen2phen.org
• The research leading to these results has
received funding from the European
Community's Seventh Framework
Programme (FP7/2007-2013) under grant
agreement number 200754 — the
GEN2PHEN project