Tweet
Hi,
Aim:
I' am comparing mpileup/bcftools and UnifiedGenotyper.
My data:
7 samples. Reads aligned with tophat against hg19.
Result:
Average 'QUAL' in vcf files: UnifiedGenotyper (404), mpileup/bcftools (28).
Average snp/indel calls: UnifiedGenotyper (49,739), mpileup/bcftools (253,933).
Issue:
Greater than 90% of the tophat-mapped reads had MAPQ of 255. Therefore, when using UnifiedGenotyper, I had to use:"-rf ReassignMappingQuality -DMQ 60".
Question:
Can I compare the 'QUAL' field in the vcf output call sets from GATK and SAMtools? My intuition says no. However, I realize GATK uses MAPQ as a cap on base quality. So if anything, the 'QUAL' values for GATK calls are low. I' am still looking into how the SNP/indel calling algorithms use MAPQ.
Any input would be helpful. I' am a novice. Also, I should maybe redo the tophat alignment for correct MAPQ.
Best.
Aim:
I' am comparing mpileup/bcftools and UnifiedGenotyper.
My data:
7 samples. Reads aligned with tophat against hg19.
Result:
Average 'QUAL' in vcf files: UnifiedGenotyper (404), mpileup/bcftools (28).
Average snp/indel calls: UnifiedGenotyper (49,739), mpileup/bcftools (253,933).
Issue:
Greater than 90% of the tophat-mapped reads had MAPQ of 255. Therefore, when using UnifiedGenotyper, I had to use:"-rf ReassignMappingQuality -DMQ 60".
Question:
Can I compare the 'QUAL' field in the vcf output call sets from GATK and SAMtools? My intuition says no. However, I realize GATK uses MAPQ as a cap on base quality. So if anything, the 'QUAL' values for GATK calls are low. I' am still looking into how the SNP/indel calling algorithms use MAPQ.
Any input would be helpful. I' am a novice. Also, I should maybe redo the tophat alignment for correct MAPQ.
Best.