find siblings by comparing vcf files

Question

[slama.slama]

Hello

When I got the results of WES in vcf format, it showed that two patients, who are supposed to be sibling, are not. I used plink for IBS/IBD estimation but didn't know how to do it (first time using plink or any other tool ). Could you please help me find a way to double-check the results. Thank you

[JC]

Not plink but time ago we published relationship computing from NGS data here:

• Relationship Estimation from Whole-Genome Sequence Data
• Accurate and Robust Prediction of Genetic Relationship from Whole-Genome Sequences

[Carambakaracho]

pardon my ignorance, but in my opinion you're conclusion might be wrong in case it's based on the two individuals only, unless you compare to their respective parents, too. Two siblings can share anything between 100% and 0% of the same genetic material. The only given is 50% shared DNA to each parent.

Or am I wrong there?

[Ram]

Any evidence one may get from using IBS/IBD scores would be inconclusive at best owing to the range of variation, as you rightly point out. Identical twins are siblings as well, and will share the Z0, Z1, Z2 values with duplicated samples. It is not possible to conclusively decide that two samples are siblings from SNV data without access to the parental SNVs as well.

[Friederike]

You fail to mention a couple of crucial details, for example:

1. Why do you think the two patients aren't siblings?
2. What was the command you used for PLINK?

Please provide these information otherwise it's difficult to help you effectively.

Answer 1

[chrchang523]

After converting the two individual-sample VCFs to plink-format, you need to merge them with "plink --bmerge", and then use plink 2.0's --make-king-table to estimate relatedness. plink 1.x --genome will NOT be effective here without more trouble than it's worth.

[slama.slama]

After converting vcf to plink format I used the following command to merge

plink --bfile data1 --bmerge data2.bed data2.bim data2.fam --make-bed --out merge

As a result, I had a FAM file and a MISSNP file.

[chrchang523]

See https://www.cog-genomics.org/plink/1.9/data#merge3 for merge-failure documentation.

Answer 2

[slama.slama]

Hello

So I got a mail with the WES result. they mentioned that 2 samples, who are supposed to be twins, are not. so they put the results in different VCF files.
knowing that it is not a family-based study, I don't know on what basis they said that.

I tried the following command to calculate IBS/IBD for the other siblings (in the same vcf file) and it works plink –vcf x.vcf --vcf-half-call m --genome --make-bed --out plinkfile

it only works when there is more than one patient in the same vcf file. so how can I calculate IBS/IBD of two or more vcf files?

Sorry if I was not very clear

thank you

[Ram]

Please edit your question and add these details in there. This content does not belong as an answer.

Once you've edited the top level post, you can delete this answer from the moderate menu.