Vcf File And Interpretation Of Allelic Depth

Question

[C00101]

I have an annotation file in VCF format for an insertion at a particular position on a human chromosome; my inquiry focuses on interpreting some of following terms in determining/indicating how real/probable a variant is from sequencing before confirming with assays. (I apologize in advance for any misconceptions that I have in using any terms.)

GT:GQ:DP:PL:AD (order of terms in file)

where GT = genotype, GQ = genotype quality, DP = read depth, PL = phred-scaled genotype likelihoods, AD = allelic depth

For example, suppose that my file has the following entry (with the organization as described above) for an individual:

0/1:99:47:624,0,503:26,21

For AD, if there is not a great disparity between the reference and alternate values (last two numbers), doe this indicate that the variant can really exist for the particular person or is more "real"?

Thanks for any help.

Answer 1

[Gabriel R.]

This field is more of help for heterozygous positions. If you have coverage 50, and you have AD 45,5, then this is really skewed towards the reference and you do not have a nice balance between ref and alt. The distribution should follow a binomial distribution. If your coverage is low, then you have less samples for the distribution and you are less able to make a call as to the reliability of het. sites.

[C00101]

Thank you for answering.

[Gabriel R.]

click on accept answer if you found the answer satisfactory :-)