Skip to content
GitLab

Chromosomal Abnormalities · Changes

Page history
Update Chromosomal Abnormalities authored by Neethu Shah's avatar Neethu Shah
Show whitespace changes
Inline Side-by-side
Chromosomal-Abnormalities.md
View page @ 2c5b27f1
Test ## Genome Variation - Chromosomal Aberrations & Others
\ No newline at end of file ​
### Context of this Interim/Proposed Descriptors
​
* Aim is to have descriptors that cover additional human genome variants, in this case large chromosomal abverrations, to support imminent new curation activities by the ACI group.
* By having proposed descriptors that clearly/uniqueky cover such variants, the ACI members can fill in the agreed upon descriptor syntax now, even in advance of CAR extension support for these.
* Furthermore, as "proposed" descriptors the descriptors documented below _may change to a very different form_ when fully implemented in the registry. In particular, ones that are weakly coordinate based may become overtly coordinate based for registration.
* However, because they cover generic classes of aberration and are specific to those, and because they follow existing (CNV) and planned SV descriptor conventions, we anticipate any syntax changes will be easily transformable via data migration code and not require user effort to convert.
- It is also possible that the registry may accept these interim syntaxes and itself convert to its preferred form.
* !!! NOTE: !!! To emphasize that last point, these descriptors should not be considered syntax that the registry will exactly support.
- These do not set a precedent or specification for the registry.
- But they do provide guidance and descriptors to consider.
​
### Approach
​
* SV-like descriptors
* Assembly doesn't really matter for these
* Employ composition
* Based on genomic variation, not resulting condition name(s)
* `;` for fusion
* Employ composition where the variation requires
* Coordinates provided as cytobands "necessary" (accepted; possibly preferred) in some cases
- `q10` is 0-length centromere (this is a genotyping standard) from q side and can be used for arm-fusions
- Not a stainable cytoband obviously.
- Similarly `p10` is 0-length centromere from the p size and can be used for arm-fusions
- For descriptors employing bands (these are large scale chromosomal abberrations after all) _other than the `p10/q10` special case_, we may translate them to genomic coordinates based on band definitions.
. Or we may decide the arm-based definitions are the "preferred" variant syntax for this scale. Must decide.
* Currently, described as variation to the haploid set.
- This raises questions for aneuploidies of sex chromosomes, since it is not "context free" and any sydromes will depend on set of _non-variant_ chromosomes (additional X is a great example of "not context free" variation)
### Aneuploidy
Easy
​
#### Missing chr
* Adapted from SV descriptor syntax:
- `GRCh38 (chr8:del)`
- `GRCh38 (chrX:del)` - Turner Syndrome (one kind)
​
#### Additional chr
* Adapted from SV descriptor syntax for `dup` (don't like it):
- `GRCh38 (chr8)x2`
- `GRCh38 (chrY)x2`
* Alt descriptor for this aneupoidy (better? see Missing chr above):
- `GRCh38 (chr8:add)`
- `GRCh38 (chrX:add)` - additional copy of X to the haploid set; Klinefelter Syndrome if male, trisomy X if female (hmm, variation is not context free)
- `GRCh38 (chrY:add)` - additional copy of Y to the haploid set; (XYY syndrome if male; if "female" actually is Klinefletter syndrom
- `GRCh38 (chr21:add)` - additional copy of 21 to the haploid set; simple trisomy 21 obviously)
* **A concern**: the variant is expressed for the haploid set, as normal. We are registering variants.
- However, the condition will be dependent on patient's diploid set.
- i.e. clinical actionability and even phenotype will depend on more than just presence of a genomic variant.
- Currently, such context considerations is not the realm of variant registry, rather it is exactly the kind of consideration the ACI effort would be concerned about: additional context of the patient (age, but also sex), syndrome risks, and various Outcome-Interventions.
- Thus, so far, we are convoluting variant descriptors with patient context.
​
#### Combine multiple, including with other SV:
​
* `GRCh38 (chr8:del),(chrY:add)`
* `GRCh38 (chrX:del),(chr1:2811131-2812146inv),(chr1:2811953-2812146)x2`
​
### Robertsonian translocation
W.L.O.G.
​
From descriptor of a human genomic variant, must consider the balanced healthy (as far as we know now) carrier from the variation that can lead to another kind of Down Syndrome
​
* `GRCh38 (chr21q;chr14q10),(chr21:del),(chr14:del)` - Look, it's simple: fusion of long arms of 21 and 14 at their centromeres, combined with loss of normal chr21 and normal chr14 in the haploid set to complete the Robertsonian translocation for a _balanced_ carrier.
* `GRCh38 (chr21q;chr14q10),(chr21p10;chr14p10),(chr21:del),(chr14:del)` - Here the balanced carrier has ALSO retained the fusion fragment of the two p arms [rare, mainly for illustration here]
* `GRCh38 (chr21q10;chr14q10)` - But the `:del` are missing, meaning the haploid set has both chr21 and chr14 as well!
- UNbalanced case.
- Down Syndrom due to effective [partial] trisomy 21 actually caused by the UNbalanced Robertsonian translocation
- Also considered more explicit: `GRCh38 (chr21q10;chr14q10):add` for this additional case
* This is a class of arm-fusion that is covered this way. Robertsonian is just one case.
​
### Others
​
#### Isochromosome
​
* `GRCh38 (chrXq10;chrXq10),(chrXp:del)` - This one causes a form of Turner syndrome due to the _partial_ loss of the p-arm of chrX.
- The isochromosome `(chrXq10;chrXq10)` is two chrX q arms fused at the centromere. The p arms are lost [causing a form of Turner]
- A class. This is just an example that is connected to Turner syndrom as well as `GRCh38 (chrX:del)`
​
#### Ring chromosome
​
* `GRCh38 (chr14p;chr14q)` - Ring chromosome. The end of chr14 p arm is fused to the end of chr14 q arm.
- We need no further banding information for this since the other end is the p10 / q10 0-length "band" used for centromeric fusions.
- This ring chr14 is likely pathogenic for Epilepsy and mental retardation. Many ring chr have been encountered, with pathogenicities; although rare.
​
### Not Actual Large-Scale Chromosomal Abberrations
​
#### Pathologic Repeats
Like Fragile X syndrome (a CGG repeat in 5' UTR of a certain chrX gene)
Express as a human genome variation.
* In fact, many of these are already present in core CAR with CAids.
- Many of these https://genome.ucsc.edu/cgi-bin/hgc?hgsid=1586229845_vq7vMKDv7xneulmCWPuaUpQGa5je&db=hg38&c=chrX&l=147912009&r=147912238&o=147912048&t=147912050&g=clinvarMain&i=CGG ClinVar ids can be found in CAR. Not all.
* Because it doesn't support cardinality in storage, CAR expands to an indel sequence (insert obviously). It has the clinvar for many Fragile X _premutatons_ (like 18 repeat copies, 37 repeat copies etc)
* But for Fragile X syndrome, problems arise with >200 repeat copies.
- These are not all in core CAR, even though the 201 ClinVar variant could be since expanded sequence is <10,000nt
- Doesn't work generically anyway for more copies or slightly longer repeat sequence.
* For which we have SV, specifically CNVs
- Express the repeat as a CNV.
- Do have to locate it. https://genome.ucsc.edu/cgi-bin/hgTracks?db=hg38&lastVirtModeType=default&lastVirtModeExtraState=&virtModeType=default&virtMode=0&nonVirtPosition=&position=chrX%3A147912048%2D147912074&hgsid=1586229845_vq7vMKDv7xneulmCWPuaUpQGa5je
​
* `GRCh38 (chrX:147912050-147912052)x201` - Fragile X with likely pathogenic phenotype exhibited; https://www.ncbi.nlm.nih.gov/clinvar/variation/183387/?new_evidence=true
* `GRCh38 (chrX:147912050-147912052)x201,500` - Covering many/all observered likely pathogenic Fragile X CNV variants
​
#### Structural Variants
While some can be large scale, can be expresseds as regular SV using genomic coordinates.
* https://genboree.org/genboreeKB/projects/clingen-allele-registry/wiki/Observations_from_Gnomad_VCF#D1-Obvious-SV-Cases-38-Syntax
* Translocations.
* Inversions.
* Large Deletions.
* Composite/Comples (inversion preceded by a duplication and followed by a deletion dupInvDel)
* Etc
* Note: The specific microdeletion(s) resulting in _Velocardiofacial syndrome_ should be provided as regular coordinate-based SV; the oft-described vague "microdeletions within a band of the 22 q-arm" is not scientific. Deletion of the entire 22q11.2 band is supported via equivalent genomic coordinate SV (deletion SV preferred over x0 CNV)
\ No newline at end of file