confusion Base quality score recalibration

hi there
i need to know about --solid_recal_mode : SET_Q_ZERO
\ --solid_nocall_strategy : LEAVE_READ_UNRECALIBRATED
i have problem in understanding these parameter though i read this explanation
--solid_recal_mode : SET_Q_ZERO
Set reference inserted bases and the previous base (because of color space alignment details) to Q0. This is the default option
When a no call is detected throw an exception to alert the user that recalibrating this SOLiD data is unsafe. This is the default option.
LEAVE_READ_UNRECALIBRATED
can you explain in simple way

Best Answers

Answers

  • h_asifh_asif Member

    Thank you so much you make very clear. one more thing what useful info i can extract from output table about covariates in tabulated form
    i am trying very hard for the graph but it is giving error (see attached) I read several users having that problem but couldnot find satisfactory solution as reported by many users.when i run VQSR there i get plots.
    I tried to fix that R problem but doesnt seem to work

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie admin

    Ah, it looks like you're missing a library. See this line in your output:

    Error in library(gplots) : there is no package called ëgplotsí
    

    You need to check whether all the dependencies are installed. Some packages are used for BQSR plotting but not for VQSR plotting, so that would explain why you can get some plots but not others.

  • h_asifh_asif Member

    you are right but i have checked R folder see print screen attached what i have in R

  • h_asifh_asif Member

    ok thanks so you mean i need to install gplots now

  • h_asifh_asif Member
    edited January 2014

    it worked Thank you so much

  • h_asifh_asif Member

    output of BQSR
    how to interpret result any link for the guide

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie admin

    Have a look at the presentation videos from our last workshop: http://www.broadinstitute.org/gatk/guide/events?id=3391

  • I am using GenomeAnalysis-3.1-1/GenomeAnalysisTK.jar -T BaseRecalibrator --defaultBaseQualities 1; is it okay ?
    My samples are from PE101 Illumina of Cancer Cell lines with no normal control.
    My questions are: --defaultBaseQualities 1
    1. can I manually filter out low quality bases later ?
    2. I think, using --defaultBaseQualities 1 will keep all bad/low quality bases and I can mannualy filter them out later ...
    May someone please guide me...
    thank you

    @h_asif said:
    hi there
    i need to know about --solid_recal_mode : SET_Q_ZERO
    \ --solid_nocall_strategy : LEAVE_READ_UNRECALIBRATED
    i have problem in understanding these parameter though i read this explanation
    --solid_recal_mode : SET_Q_ZERO
    Set reference inserted bases and the previous base (because of color space alignment details) to Q0. This is the default option
    When a no call is detected throw an exception to alert the user that recalibrating this SOLiD data is unsafe. This is the default option.
    LEAVE_READ_UNRECALIBRATED
    can you explain in simple way
    @Geraldine_VdAuwera said:
    Hi there,

    These parameters allow you to deal with the specific problems that can arise with SOLiD data, particularly the problem of reference inserted bases. By default the GATK will mark these bases as Q0 (=no base call), which means they cannot be trusted. This is the safest option. The second option determines how the program will deal with Q0/ no-call bases in the second step. By default the GATK will quit with an error when it finds such a base, to tell you that the data has problems and it cannot be safely recalibrated. If you set this to LEAVE_READ_UNRECALIBRATED, the GATK will not modify the read. This may cause problems in later analyses. If you set it to PURGE_READS, the reads will be marked as bad and will not be used in later analyses. This option is safe to use but it does mean you will lose some data. If there are not many no-calls it is fine, but if there are many it could be a problem.

    I hope this helps, good luck.

  • Geraldine_VdAuweraGeraldine_VdAuwera Cambridge, MAMember, Administrator, Broadie admin

    @charitrakumarmishra‌, if you're working with Illumina data you don't need to set this argument. You should certainly not use it like this. Just use the default arguments; low quality bases will be treated appropriately during variant calling.

Sign In or Register to comment.