calibration with both Novoalign and GATK ?

[clicker]

Hi,

I'm new to sequence analysis. I've managed to do alignments with novoalign, removed low mapping reads, and removed presumed PCR duplicates, and then done variant calling with bcftools/vcfutils.

I've now understood I might need to recalibrate my basecalls, and I can do this with either GATK or novoalign (during the alignment step).

I'm assuming I can't do both - as the effects of one would get amplified by the effects of the other. Or is this incorrect?

Also, given I'm past the alignment stage now - and filtered out PCR duplicates and low mapping reads - can I simply run GATK now and assume it will fix up whatever is remaining? Or should I run GATK before I remove the duplicates and low mapping reads (i.e. will the information from them be used in the calibration)?

Thank you.

Tom

[sparks]

Hi Tom,

You can run Novoalign and GATK quality calibration without any problems. Novoalign quality calibration happens during alignment and can improve sensitivity and specificity of the mappings. GATK is post alignment. Both will improve accuracy of SNP calls.

Colin