Nanopore reads begin and end in an A polymer region of Arabidopsis but only in one direction?

[dbcrosbie]

I am assembling genomes of several A. thaliana accessions sequenced with Nanopore and have noticed something strange. After realigning the reads to the assembly we see a large amount of forward reads will end and begin in some regions of 3 or more A, however there are some reads that span these regions and the reverse reads do not have this problem at all. The image provided shows the reverse reads (red) relatively stable in this region but the forward (green) show a steep kink. Any idea if this is an alignment, assembly or sequencing quirk?

[GenomicSeq]

dbcrosbie The behavior you're observing in your sequencing is interesting. But I think it's hard to determine the exact cause without examining the raw data. It's possible that there is a sequencing bias with the Nanopore technology, specifically with homopolymer regions. I've heard of nanopore sequencing having issues with accurately sequencing homopolymeric regions due to how the technology works. This might lead to errors in basecalling or misalignments, especially in regions with repetitive sequences.


Also, it's possible that the alignment tool and its settings could be influencing this. Some aligners might be more tolerant to homopolymers and other sequence nuances of Nanopore data. It could be worthwhile to test multiple alignment tools or tweak parameters and see if this continues to happen.