MarkDuplicates RG ID on SAMRecord not found in header

Question

[bruce.moran]

I am calling variants in RNAseq and exome data. Aligning with STAR, BWA MEM respectively. Get to MarkDuplicates and STAR BAMs give:

Ignoring SAM validation error: ERROR: Record 1, Read name HWI-ST968:151:C1MMAACXX:1:1112:9825:81904, RG ID on SAMRecord not found in header: XYZ  PL:ILLUMINA     SM:PQR        DS:123       CN:NOP       LB:LANE_X       DT:2018-10-01T20:28:04

But looking at this read there is a valid RG:Z field

HWI-ST968:151:C1MMAACXX:1:1112:9825:81904       163     1       27074   60      23S68M2S        =       27089   68      CCAGCATTCCCGCCCGGAAAACTCCACGGACAGAAGAGCCCGGCCGGCCACAGTCCATGGGGTCTCAAAGAGTGGGACATGACTGAGTGACCA       CCCFFFFFHHHHHJJJJJJJJJGIGJIJIIJIJIJHIJJJJIEHHFDDDD@AADCCADDDDDDDDDEDDDDD>CBDDDDDDCDDCCD@CCCCD       NH:i:1  HI:i:1  AS:i:103        nM:i:8  NM:i:5  MD:Z:3T7G5T1A30C17  jM:B:c,-1       jI:B:i,-1       rB:B:i,24,91,27074,27141        MC:Z:53M40S     RG:Z:XYZ      PL:ILLUMINA   SM:PQR DS:33_L2        CN:NOP       LB:LANE_X       DT:2018-10-01T20:28:04

And the header of the BAM, specified by --outSAMattrRGline "$RGLINE" includes: @RG ID:XYZ PL:ILLUMINA SM:PQR DS:123 CN:NOP LB:LANE_X DT:2018-10-01T20:28:04

Answer 1

[Pierre Lindenbaum]

yes there is RG:Z:XYZ in the read, but as the message says:

 not found in header

I'm pretty sure that there is no such read group

@RG ID:XYZ

in the bam header. Check your upstream workflow of analysis.

[bruce.moran]

Thanks for the input. As soon as I posted I found the issue, as usual. I did have an @RG header, and it <looked> correct, but I used a tab-delimited variable via RGLINE=$(echo -e "ID:XYZ\tPL:ILLUMINA"), whereas STAR wants it as space-separated list. ValidateSamFile is happy with it anyway.