Found contigs with the same name but different lengths; BQSR

SystemSystem Administrator admin
This discussion was created from comments split from: Does the Broad still recommend aligning WXS/WES data to b37/hg19?.

Comments

  • manbamanba Member

    thanks, is hg38 ok for production. 1KG and ucsc also provide the reference, is there some difference and application restrcitions for use.

    I ran with an error when do BQSR command, it said "

    A USER ERROR has occurred: Input files reference and features have incompatible contigs: Found contigs with the same name but different lengths:
    contig reference = chr1 / 248956422
    contig features = chr1 / 249250621.

  • manbamanba Member
    liftover is not a easy job, the cram data in 1KG make me upset, do you have some experience in converting them, and finally change to hg19 version, I ran with a lot of errors
  • bhanuGandhambhanuGandham Member, Administrator, Broadie, Moderator admin

    Hi @manba

    Then I suggest you use the same reference genome build(hg38) for the known sites in BQSR, as that is the same reference that was used for alignment.

  • manbamanba Member

    @bhanuGandham said:
    Hi @manba

    Then I suggest you use the same reference genome build(hg38) for the known sites in BQSR, as that is the same reference that was used for alignment.

    to be honest, I really use hg38, in your bundle hg38 version and 1000 genome hg38 version, my bam was converted from cram in 1000 genome website of exome data, really feeling said

  • SkyWarriorSkyWarrior TurkeyMember ✭✭✭

    This is a simple CRLF error. There is one additional line ending character at the end of each contig.

  • manbamanba Member

    @SkyWarrior said:
    This is a simple CRLF error. There is one additional line ending character at the end of each contig.

    Thanks for your kind help
    I am not quite understand your meaning, can you clarify more clearly, if there can be an example, will be better,

  • manbamanba Member

    In fact, I made a bed file from the reference .fai file, and use samtools view -L xx.bed input.bam > output.bam. to make each chomosome LN be the same as reference, but still gave me the same error

  • bhanuGandhambhanuGandham Member, Administrator, Broadie, Moderator admin

    HI @manba

    Please refer to this thread for more information on CRLF error @SkyWarrior mentioned. @SkyWarrior please feel free to share your two cents on this if you have seen such an error before.

    Regards
    Bhanu

  • manbamanba Member


    contig may not have additional line ending

  • SkyWarriorSkyWarrior TurkeyMember ✭✭✭
    edited December 5

    HUGE EDIT:

    I have not paid enough attention to the contig lengths. The reference that you are using is HG38 however the reference that this bam file is mapped to is hg19.

    Check these 2 dict files

    Post edited by SkyWarrior on
Sign In or Register to comment.