I have found SNP on my datasets using SAM tools mpileup, GATK abd Freebayes software. I can see different SNP quality for the same SNP in each tool.
I would like to know what should be the minimum SNP quality score to consider it as SNP based on the quality score?
Well...
This question is not only about the quality value, but how you processed your data before you do the calling and the quality of your sample!
I don't know how you did the QC before it, but based at the common used pipeline of GATK you did the recalibration of the bases. Different from the GATK, at mpileup this step doesnt exist so it brings you the "machine quality" as a parameter to calculate the QUAL value (probably being a lil higher than GATK QUAL value). You should also pay attention at the parameters used to do the calling (-q at samtools view step, and -stand_call/emit at GATK). Are you considering the Phred Q of the bases too low?
It's hard to say "use this QUAL value"... there are many variables in this equation... but based at GATK research forum, don't trust Q value lower than 30... I would suggest 60 as a standard based in my analysis, but you need fully understand your data before you try a Q value to base yourself.
- Badges
- Science method