running on frameshift corrected contigs

[chiaramazzoni]

Hi! I'm working with metagenomics data, I have both long and short reads (single end) at my disposal, and before finding your tool my plan was assembling (with metaflye) and then binning (considering different tools). Now with LRbinner I'm debating over two options:

1. using long reads (+ short reads -based contigs) in reads mode and assemble later;
2. using assembled long reads (+short reads -based contigs) in contigs mode.
   The question with option 2. is that I performed frameshift correction of the contigs using a translated search against a protein reference database. Corrected sequences will either have 'N' or 'NN'. Will these be detected as part of relevant 3-mers or 4-mers?
   Thank you in advance!

[anuradhawick]

I am skipping bases having N in k-mer counting. Hopefully, it will be okay. Let me know if you face anything unexpected.