readgroup questions

[mrfox]

hi all,
I know that there have been many posts about readgroup in SAM/BAM format. But I am still confused.

Assume I have a sample named "Treated"sequenced with Illumina HiSeq platform. There are 5 pairs of the fastq files. I map each of the small fastq file using BWA and add read group when I merge the pairs of small bam files. For the first pair of bam files, can I just use "Treated_1" as its readgroup and "Treated_2" for the 2nd pair of bam files?

An opposite question: if I merge the fastq files for one end to get a file s1.fastq, then I merge the fastq files for the other end to get s2.fastq. Then I removed the original 5 pairs of original small fastq files. My question is, is there any to for me to seperate the merged s1.fq and s2.fq back into the 5 pairs of small fastq files?

Hope my description is clear. Thank you all!

[TiborNagy]

It is depends how do you run bwa. If you use bwa samse for each fastq files, then the readgroup can identify the pairs individually (you get 10 different readgroups). If you use bwa sampe, then readgroups identify the samples (you get 5 different readgroups).

No. If you merge your fastq files before alignment, you can not separate them from the bam file.

[mrfox]

Thank you TiborNagy.

I agree with you. Unfortunately, I mapped the merged fastq files and the original small fastq files were deleted accidentally. I realized that properly adding readgroup info from the bam file is impossible. I will request the original small fastq files and re-map them.