Tweet
I have a non-barcoded multiple samples pool of long amplicons sequenced on RSII. I'm trying to figure out the correct workflow. My goal is to perform variant calling.
After reading on this for 2 days this is what I have pieced together:
1. bax2bam conversion
2. blasr alignment all reads to my reference
3. quiver/arrow polishing
4. LAA generate consensus reads
5. blasr (or bowtie even?) to align consensus reads to reference
6. minorseq for variant calling
Does this look correct?
I initially thought CCS subreads would be a good trick to demultiplex my multiple samples (about 70? I don't know why it wasn't barcoded in the first place, I didn't handle that part). However my amplicons are all >5kb and I think I read CCS subreads are more useful for small amplicons (makes sense if these are able to run longer). I just want to see the extent of variation (primarily small variation) in this population of 70 samples.
After reading on this for 2 days this is what I have pieced together:
1. bax2bam conversion
2. blasr alignment all reads to my reference
3. quiver/arrow polishing
4. LAA generate consensus reads
5. blasr (or bowtie even?) to align consensus reads to reference
6. minorseq for variant calling
Does this look correct?
I initially thought CCS subreads would be a good trick to demultiplex my multiple samples (about 70? I don't know why it wasn't barcoded in the first place, I didn't handle that part). However my amplicons are all >5kb and I think I read CCS subreads are more useful for small amplicons (makes sense if these are able to run longer). I just want to see the extent of variation (primarily small variation) in this population of 70 samples.