using vcffilter

Question

[Aladdin]

hello, so we are doing this tutorial in bioinformatics, and I am asked to use a command "vcffilter" to filter the results of variant calling in 2 strains, I will post the task's question, I tried a command but it ended up deleting everything in my .vcf file. the command i used was:

vcffilter -f "MQ" "QUAL" "DP" AK83_var.vcf > AK83_var_filtered.vcf

this didnt work. the script of the task: "The detected variants might be either the property of an isolate or a result of some technical errors. Among other factors, sequencing errors, ambiguous mapping and PCR duplicates cause calling of false-positive variants. Therefore, additional filtering needs to be applied to distinguish between correctly inferred variants and variant “noise”. Filter the results by mapping quality (MQ), quality (QUAL), and read depth (DP) using vcffilter (you may also use any other filter parameters or combination of parameters)."

Answer 1

[Giacomo Rocchetti]

First of all, you should understand better the objective of the task. In particular, what does Filter the results by mapping quality (MQ), quality (QUAL), and read depth (DP) mean? With which values should I consider a VCF record valid?
For example, should I keep a variant with a DP of 4? Is a MQ of 40 enough?

With that being said, the command you are using is missing the thresholds for each filter condition. Here is an extract of the vcffilter's documentation:

Filters are specified in the form "<ID> <operator> <value>:

• -f "DP > 10" # for info fields
• -g "GT = 1|1" # for genotype fields
• -f "CpG" # for 'flag' fields

Hope this helps.

[Aladdin]

Aha i see, I will consider this and give it a run, thank you

[Aladdin]

okay so I checked my .vcf file and its immense data, however, the MQ is mostlly at a value of 60, there are values of 42, 38... with that being said, should I filter those values of 42, 38...

[Giacomo Rocchetti]

There is no general answer, it depends on your use case. Of course if you want only perfectly mapped variants, then set the threshold to 60.
Considering that the task you are following does not specify any threshold, I can't give you any insight.