Plasmid study using metaSPAdes, after having analysed host with MEGAHIT assemblies

[hellopeccat]

Hi there,

I noticed that the k-mer of 141 used in MEGAHIT would not be compatible with the request of metaplasmidSPAdes in Discussion#998.

Since I have finished the analyses of host communities, based on the MEGAHIT assemblies, I would like to get some ideas or suggestions from the forum family. My concerns are:

1. Is it an acceptable idea to analyse host and plasmid separately using different assembler tools, and finally combine the results in one study?
2. If that idea is stupid, given that I have no time to re-analyse the host, would you suggest some alternative way for plasmid recovery in the host-plasmid study? For example, how about MEGAHIT + SCAPP?

Many thanks in advance. Also, thanks to the excellent work of SPAdes.