lh3/fermi-lite

considerations for local assembly

Opened this issue · 1 comments

ekg commented

We would like to deploy fermi-lite in vg for local assembly and homogenization. Is there any particular consideration that we should take when doing this?

It may be helpful to assemble the data from many genomes in a small region (1kb-100kb for instance). What parameters might we use in that case?

lh3 commented

Although fermi-lite is supposed to do local reassembly for human, I have never tried it for this particular use case. I was mostly using the algorithm for whole-genome assembly. Not sure how much attention I need to pay to local assembly. I have not tried to assemble many genomes, either. Let me if you see anything unusual.