This is a test version of Biostars. For the public version, visit https://www.biostars.org.
Query a single sample VCF for a list of rsid, do calculations and generate a personal genomics report

Hi, I need to generate a personal genomics report based on some rsid (mainly SNPs), starting from single sample VCFs. Till now, I manually search for the genotype (0/0, for example) of every rsid,I copy the value in an excel file, I generate the corresponded genotype (AA, for example) and i trigger some rules to generate some results (for example: if genotype is AA: low risk, if genotype is AC increased risk...). Now, I want otop create some scripts to query the VCF for a list of rsid, have the correspondent genotypes, do some calculations like in excel, then generate an output file that I can switch to a readable report.

I seee there are solutions like scikit + numpy or pèandas, or similar, or, to convert the VCF in Parquet and then use cloud solutions in Google Cloud, Amazon, or Azure...but I not know if this is the better approach....

Do you have some ideas? Thanks!

personal query genomics vcf
• 1,512 views • link
written by nico77 • nico77

1 answer

Now, I want otop create some scripts to query the VCF for a list of rsid

https://samtools.github.io/bcftools/bcftools.html#expressions

bcftools view -i 'ID=@filelistofrsid.txt' input.vcf

have the correspondent genotypes

https://samtools.github.io/bcftools/bcftools.html#query

bcftools query -f [%ID %TGT \n];

do some calculations like in excel, then generate an output file that I can switch to a readable report.

awk https://www.google.com/search?q=awk+tutorial

• 0 views • link
written by Pierre Lindenbaum Pierre Lindenbaum

Thank you Pierre.

So, I have to create a text file with the list of RSID I want to query (filelistofrsid.txt), right? Then bcftools will search for the correspondent genotypes in my VCF (input.vcf), right? And the output would be what kind of file? Thanks!

• 0 views • link
written by nico77 • nico77

Log in to answer this question.

More posts like this