Compare GATK and Samtools

[LohiH]

Hi all,

I am interested in comparing the tools GATK and Samtools. Both of them use the probabilistic methods to estimate the genotypes. Could anyone suggest an ideal approach to compare the difference in the variants detected by both the algorithms?

Regards
HL

[SHI Q.]

I suggest to use some public sequence data, like 1000genomics etc. and use GATK and samtools to call snps, then figure out the TP,TN,FP,FN and time cost and member. the fact there are a lot of papers did these things already.

[LohiH]

Hi

I am working with my own sequencing data to call the variants.So, I would like to compare them with the known database SNP's but which statistical parameter would be ideal to choose to compare the confidence between the two algorithms? like depth, mapping quality and quality score?

Another query regarding the vcf output. In the VCF files from both GATK and Samtools, there is a column in the output "QUAL" which is a phred-scaled probability score for the alternate allele to be wrong. But, for the same SNP, the QUAL values are different for both the algorithms. Below are the QUAL values for the two methods, could anyone explain why they differ?

Samtools-QUAL value is 222
GATK QUAL value is 1513