Genes that arose from a common ancestor by duplication, called paralogs, often
keep a similar protein sequence and three-dimensional structure long after they
have taken on different roles. Because equivalent positions in a family of
paralogs tend to play equivalent structural or functional roles, a variant known
to cause disease at one position in one gene points to a candidate
disease-associated position at the aligned residue in each of its paralogs. This
idea, known as paralogue annotation, was introduced for the cardiac ion
channel genes and has been used to help classify variants of uncertain
significance.
This track container takes every ClinVar variant that changes a protein residue
-and projects it onto the equivalent residue of each paralog of the gene it
-occurs in. It has two parts:
+and projects it onto the equivalent residue in related genes. It offers two ways
+of deciding which positions are equivalent, and shows the evidence for each.
+
+
-The mapping is purely positional: it indicates that a residue is the alignment
-equivalent of a position that is variant in a paralog. It does not itself assign
-pathogenicity, and a projection is only as reliable as the underlying alignment,
-so the percent-identity and residue-conservation fields should be used to judge
-confidence.
+The mapping is purely positional: it indicates that a residue is the equivalent,
+by alignment or by shared domain position, of a position that is variant in a
+related gene. It does not itself assign pathogenicity, and a projection is only as
+reliable as the underlying alignment, so the residue-conservation field, and for
+paralogs the percent-identity fields, should be used to judge confidence.
In the Paralog Variants track, each feature marks the codon in this gene that
aligns to a protein-changing ClinVar variant in a paralog. Features are colored
by the ClinVar clinical significance of the source variant:
| |
Pathogenic or likely pathogenic |
|---|
| |
Benign or likely benign |
|---|
| |
Uncertain significance |
|---|
@@ -57,92 +77,121 @@
The details page for each mapped variant links to the original ClinVar record
and to the position of the source variant on the genome. The track can be
filtered by clinical significance, ClinVar review status (0 to 4 stars),
molecular consequence, residue conservation (identical, similar or different
between the two genes), Ensembl paralog percent identity, and the percent
identity of the pairwise alignment. Because most protein-changing ClinVar
variants are classified as uncertain, filtering to pathogenic and likely
pathogenic classifications and to higher alignment identities is recommended for
variant-interpretation use.
The Paralog Alignments track shows each pairwise protein alignment as a
blocked feature; the label is the partner gene, and the aligned blocks mark the
regions that could be compared between the two proteins.
+
+The Pfam Domain Variants track uses the same clinical-significance colors. Each
+feature marks the codon in this gene that shares a domain position with a
+protein-changing ClinVar variant in another gene, and its details give the domain,
+the domain position, the source gene and variant, and whether the residue is the
+same as in the source gene. It can be filtered by source gene, Pfam domain,
+clinical significance, review status, molecular consequence and residue
+conservation. The Pfam Domains track shows the domain footprints, labeled and
+searchable by domain name and gene.
+
Methods
Paralog relationships and their percent identities were taken from Ensembl
BioMart (release 116) as within-species paralog pairs between Ensembl genes. For
each protein-coding gene, the single representative transcript and protein were
taken from the MANE Select set, so that residue numbering matches the transcripts
used in clinical variant reporting. Each pair of paralogs whose proteins share at
least 20% identity was aligned with a global Needleman-Wunsch alignment
(BLOSUM62), and the aligned residue positions were recorded. Pairs below 20%
identity were not aligned, because alignments in that range are not reliable
enough to transfer positions.
Every ClinVar variant with a protein-residue consequence (missense, stop gain or
loss, start loss and in-frame insertions or deletions) was assigned to the codon
of the overlapping MANE Select transcript, and its reference residue was taken
from the translated protein. Each variant was then projected across the
alignment to the equivalent residue of every paralog, and that residue was mapped
back to its genomic codon to place the feature. Variants that align to a gap, or
that occur in a gene with no paralog, were not projected. The pairwise alignments
themselves were converted to PSL to form the Paralog Alignments track.
+For the domain method, the MANE Select proteins were searched against the Pfam-A
+profile library (release 38.2) with HMMER, using each family's curated gathering
+threshold to decide domain membership. For every domain family found in two or
+more genes, the matched protein regions were aligned to the family model, and each
+residue was assigned the model match-state column it occupies; this column is the
+shared numbering that makes a position in one gene comparable to a position in
+another. Each ClinVar coding variant that falls on a domain column was then
+carried to the residue in the same column in every other gene that has the domain,
+and that residue was mapped back to its genomic codon. A residue that falls
+between the modeled positions of a domain, in an insertion relative to the model,
+has no shared column and was not carried across. Because a few domains occur in
+hundreds of genes, each variant is shown at its 25 most conserved equivalents
+(residues identical to the source residue first, then chemically similar, then
+different); without this limit the largest families would produce hundreds of
+projections per variant. The Pfam Domains track records, for each domain instance,
+the gene, the domain and the range of model columns it covers.
+
+
The ClinVar data were taken from the UCSC ClinVar track for this assembly. The
paralog list was downloaded from the Ensembl BioMart service at
https://www.ensembl.org/biomart/martservice,
and the MANE Select set from the UCSC MANE track. The commands used to build the
track are documented in the
makeDoc,
and the processing scripts are in the
kent source tree.
Data Access
The data can be explored interactively in table format with the
Table Browser or the
Data Integrator and exported from there to spreadsheet
or tab-sep tables. From scripts, the data can be accessed through our
API, track=clinvarMappedParalog.
For automated download and analysis, the annotation is stored in bigBed files
that can be downloaded from
our download server.
-The files for this track are called clinvarMappedParalog.bb and
-clinvarMappedParalogAln.bb. Individual regions or the whole genome annotation
+The files for this track are called clinvarMappedParalog.bb,
+clinvarMappedParalogAln.bb, clinvarMappedPfam.bb and
+clinvarMappedPfamAln.bb. Individual regions or the whole genome annotation
can be obtained using our tool bigBedToBed, which can be compiled from
the source code or downloaded as a precompiled binary for your system.
Instructions for downloading source code and binaries can be found
here.
The tool can also be used to obtain features within a given range, e.g.
bigBedToBed http://hgdownload.soe.ucsc.edu/gbdb/hg38/clinvarMapped/clinvarMappedParalog.bb -chrom=chr2 -start=165000000 -end=166000000 stdout.
Credits
Thanks to Ensembl for the paralog annotations, to the MANE collaboration between
-NCBI and EMBL-EBI for the reference transcript set, and to ClinVar for the variant
-data. The paralogue annotation concept was developed by Ware, Walsh, Cook and
-colleagues.
+NCBI and EMBL-EBI for the reference transcript set, to the Pfam team at EMBL-EBI
+for the protein domain models, and to ClinVar for the variant data. The paralogue
+annotation concept was developed by Ware, Walsh, Cook and colleagues.
References
Ware JS, Walsh R, Cunningham F, Birney E, Cook SA.
Paralogous annotation of disease-causing variants in long QT syndrome genes.
Hum Mutat. 2012 Aug;33(8):1188-1191.
PMID: 22581653; PMC: PMC4640174
@@ -157,15 +206,25 @@
href="https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3888601/" target="_blank">PMC3888601
Morales J, Pujar S, Loveland JE, Astashyn A, Bennett R, Berry A, Cox E, Davidson C, Ermolaeva O,
Farrell CM et al.
A joint NCBI and EMBL-EBI transcript set for clinical genomics and research.
Nature. 2022 Apr;604(7905):310-315.
PMID: 35388217; PMC: PMC9007741
+
+Mistry J, Chuguransky S, Williams L, Qureshi M, Salazar GA, Sonnhammer ELL, Tosatto SCE, Paladin L,
+Raj S, Richardson LJ et al.
+
+Pfam: The protein families database in 2021.
+Nucleic Acids Res. 2021 Jan 8;49(D1):D412-D419.
+PMID: 33125078; PMC: PMC7779014
+
+