What is the easiest way to count the number of SNPs in humans (and distinguish them by the standard NCBI filters of "missense mutations" / "nonsense mutations" / "synonymous mutations" / "intron variants")? I wish to get the data from NCBI's dbSNP database. I can see the numbers through the NCBI variation viewer. Tried doing it through NCBI Entrez Direct but haven't gotten it to work thus far. Any ideas?
5 answers
Hello,
one way is to download the dbSnp vcf file an annotate it with snpEff (with -canon option). snpEff produces a summary file containing the infos you like.
As the dbSnp file is huge it might take some time to annotate ...
If you are just interested in a given gene, you can first build a subset of variants using tabix and annotate than.
fin swimmer
Try the following.
$ esearch -db snp -query "brca1" | esummary | xtract -pattern DocumentSummary -element SNP_ID -element FXN_CLASS -element VALIDATED
1491587915 intron-variant no-info
1491573730 intron-variant no-info
1491558507 intron-variant no-info
1491544899 frameshift-variant,nc-transcript-variant,reference,upstream-variant-2KB,utr-variant-5-prime no-info
1491538523 intron-variant no-info
1491356043 intron-variant by-cluster
1491320693 intron-variant no-info
1491293961 frameshift-variant,nc-transcript-variant,reference,upstream-variant-2KB,utr-variant-5-prime no-info
1491277616 intron-variant no-info
Truncated due to space constraint.
One more variation. If you only need numbers.
$ esearch -db snp -query "brca1" | esummary | xtract -pattern DocumentSummary -element SNP_ID -element FXN_CLASS -element VALIDATED | awk -F '\t' '{print $2}' | sort | uniq -c
2 cds-indel,frameshift-variant,intron-variant,nc-transcript-variant,reference
49 cds-indel,intron-variant,nc-transcript-variant
22 cds-indel,nc-transcript-variant
1 cds-indel,nc-transcript-variant,reference
1 cds-indel,nc-transcript-variant,upstream-variant-2KB
1 cds-indel,nc-transcript-variant,upstream-variant-2KB,utr-variant-5-prime
1 cds-indel,nc-transcript-variant,utr-variant-3-prime
1 cds-indel,splice-acceptor-variant
93 downstream-variant-500B
125 downstream-variant-500B,nc-transcript-variant,utr-variant-3-prime
1 frameshift-variant,intron-variant,nc-transcript-variant
1144 frameshift-variant,intron-variant,nc-transcript-variant,reference
10 frameshift-variant,intron-variant,reference
2 frameshift-variant,nc-transcript-variant
459 frameshift-variant,nc-transcript-variant,reference
1 frameshift-variant,nc-transcript-variant,reference,splice-donor-variant
39 frameshift-variant,nc-transcript-variant,reference,upstream-variant-2KB,utr-variant-5-prime
36 frameshift-variant,nc-transcript-variant,reference,utr-variant-3-prime
2 frameshift-variant,nc-transcript-variant,reference,utr-variant-5-prime
16573 intron-variant
86 nc-transcript-variant,reference,stop-gained
5 nc-transcript-variant,reference,stop-gained,synonymous-codon
2 nc-transcript-variant,reference,stop-gained,upstream-variant-2KB,utr-variant-5-prime
5 nc-transcript-variant,reference,stop-gained,utr-variant-3-prime
Truncated due to space constraint.
You can get this info using the myvariant.info API. For example, suppose you are interested in SYNE1 (ENSG00000131018), you can get all known variants associated with this gene using this API call:
http://myvariant.info/v1/query?q=cadd.gene.gene_id:ENSG00000131018
which currently says that there are 88731 such variants.
To count the number of missense you can execute this call and see that there are currently 56500:
http://myvariant.info/v1/query?q=cadd.gene.gene_id:ENSG00000131018%20AND%20cadd.consdetail:missense
similarly, there are 3893 "stop_gained" or nonsense variants:
You can also consult this page http://docs.myvariant.info/en/latest/doc/packages.html to learn more about the myvariant libraries for python, R, and node.
Log in to answer this question.