Samtools: Variants quality score biased to 222 and 228

[newbird]

Hi there!

I am new in Exome sequencing. I am having problem with the SNP quality score provided by SamTools mpileup. The scores are ranged from 3 to 228 and ~40% SNPs have Qual score of 222 and ~36% have 228. I do not understand why the result is not evenly distributed and having large peaks at 222 and 228. Any help would be highly appreciated.

I used GATK-3.3 with the following arguments:

1. GenomeAnalysisTK.jar -T BaseRecalibrator -nct 8 -rf BadCigar -R hg19.fa QualityScoreCovariate -cov CycleCovariate -cov ContextCovariate

2. GenomeAnalysisTK.jar -T PrintReads -nct 8 -R hg19.fa -BQSR

And, I used Samtools with the following arguments

3. samtools mpileup -C 50 -uf my.bam | bcftools call -mv -Ob > my.recal.bcf

4. bcftools view my.recal.bcf | perl vcfutils.pl varFilter -Q 25 -d 8 -D 20000 > my.recal.flt.vcf


Thanks.