Entering edit mode
4.0 years ago
ariel
▴
250
The clinvar FTP site has two files that contain variant annotations:
- "tab_delimited/variant_summary.txt.gz"
- "vcf_GRCh37/clinvar.vcf.gz"
The summary table contains variants from multiple assemblies. However, if I filter to just GRCh37, I get roughly the same number of results (rows, or total number of calls) as in the VCF file.
However, when I annotate variants in my clinical samples from these two sources, I get an order of magnitude more hits from the summary table than the VCF table.
Can someone please compare and contrast these two ClinVar sources for me?
How are you annotating variants? That might make a huge difference.
Can you show us your commands for both the table file and the VCF file?
The two files in question are
and
By annotating, I just mean joining my table to the clinvar table by genomic coordinates. But the question is really about these two files from ClinVar that both seem to have annotated variants. I don't know what the differences and similarities are, and what is the use case for each. The ClinVar website doesn't provide any insight.
Please tell us exactly how you're doing it. The devil is in the details. The two files should be equivalent except for the GRCh37/GRCh38 difference you already listed. I recently used the tab delimited file, but the VCF file should be fine too. The co-ordinates might be zero- or one-based, so watch out for that.
Perhaps I'm not being clear. Forget the fact that I'm doing anything with the tables. The essence is that ClinVar publishes two very similar tables of variants, but does not provide any documentation on how they relate.
No, ClinVar does provide documentation on the VCF. See:
Thanks! That first link does really help. Now I understand the difference.