Which elements of the read groupe header are considered in BQSR?

rburrirburri Posts: 2Member

I am working on a data set in which (1) multiple individuals were sequenced on the same lane, and (2) the same individuals were sequenced in multiple runs. If I get this right, we would thus want to consider both lane and run as covariates in BQSR. I have two questions related to this:
1) Which elements of the @RG header are considered by BQSR? All given there (e.g. ID, SM, LB, PI, PL)?
2) I am not sure where (in which @RG elemnt) to best incorporate run and lane info?



Best Answers


  • rburrirburri Posts: 2Member

    Thanks a lot Geraldine! With 'run' I was referring to the flow cell, but I see the point here. What I still wonder is whether BQSR consideres the ID field exclusively, or whether it also considers different libraries (LB in the RG header) that have been sequenced for the same individual?

  • EugenieEugenie Posts: 17Member

    It seems that to cat fastq's from different lanes together before the alingment was not the best idea...) Dear Geraldine, please, can you share if your practice is to align reads from different lanes (but for the same sample) separately and then merge n .sam's together? These two ways seemed equal to me (and even fastq's merging seems to have some advantages http://seqanswers.com/forums/showthread.php?t=19618) and I chose the easiest but now I realize that for BQSR it can make difference... I see that the question is a bit out off topic but as it interferes with the GATK application;) Thank you!

  • Geraldine_VdAuweraGeraldine_VdAuwera Posts: 9,962Administrator, Dev admin

    Ah, yes typically we process lane data (even from the same sample) individually from alignment all the way through pre-processing -- although there are some optional post-per-sample-merging steps. Have a look at these FAQ articles:


    Geraldine Van der Auwera, PhD

Sign In or Register to comment.