Maybe I find a new bug about RG option

[yshcai]

Hi, I run the ngmlr followed command --ngmlr -t 10 -r $ref -q $reads --rg-id Pse2Mcin --rg-sm Pse -o out.sam. After running done, I use the samtools--v1.9 to convert sam file to bam file, and sort the bam file. The command like this: samtools view -@ 10 -bS out.sam > out.bam && samtools sort -@ 10 -o out.sorted.bam -T tmp.ali out.bam. When I use the out.sorted.bam as the input of PBSV (SV caller), the log file shows BAM header: read group ID not found. So I check the orignal out.sam generated by the ngmlr, I find there are some weired infomation. First, I check whether the out.sam file contains the RG line or not, and the result is @RG ID:Pse2Mcin SM:Pse. So I think that's right. Second, I check whether all read alignments contains the RG:Z:Pse2Mcin or not. And the result is surprised to me, there are some line contain the RG:Z:Pse2Mcin, but other lines don't. I select two alignments as examples below.

m64399e_220420_071234/271/ccs   0       ptg000008l_1    2331517 58      2S2853M1D3463M1D6472M   *       0       0       RG:Z:Pse2Mcin   AS:i:25566      NM:i:2  XI:f:0.9998     XS:i:0  XE:i:25566      XR:i:12788      MD:Z:2853^T3463^A6472   SV:i:2  QS:i:2  QE:i:12790      CV:f:99.984360  
m64399e_220420_071234/296/ccs   0       ptg000002l_1    70091   59      1S11663M1D1092M1D223M1I2568M1I178M1D1787M4S     *       0       0       AS:i:34997      NM:i:5  XI:f:0.9997     XS:i:0  XE:i:34997      XR:i:17513      MD:Z:11663^T1092^A2969^C1787    SV:i:2  QS:i:1  QE:i:17514      CV:f:99.971458  

Although I don't know what cause this case, I continue running PBSV to call structural variations by adding the option --sample which can override sample name tag from BAM read group. And it does work without any error, but I don't know this is correct or not.
I'm looking forward to your reply! Thank you!