hi foks,
Not too surprisingly, I'm seeing several benefits of using a 300PE approach for metagenomic (and isolate) sequencing, even without tuning the target insert size to fully leverage the longer read lengths. I'm now in the process of extending and optimizing insert sizes for this approach and wanted to source thoughts from the community on part of the computational side. I typically assemble Illumina data with Spades and, as mentioned here, Spades can be rebuilt to accept kmers > 127. For high quality 300PE reads, what are folks thoughts on min and max useful kmer lengths? I don't suppose kmers much > 251 or so would provide a tangible benefit (though I'd love to hear otherwise), but also wondering what min kmer length seems useful? Eg. is there likely much benefit to kmer length of 21, 33, 55 for this approach? All thoughts or data welcome!
thanks!
