Coverage calculation prior to a genome assembly-BBTools

Question

[GLR]

Hello,

I'm currently following the BBMAP pre-processing guide prior to performing a genome assembly and it mentions normalisation and subsampling for assemblies with high or uneven coverage. Having done a bit of reading about calculating coverage I was wondering which is best way to go about it?

1) Map reads to a reference genome (reference genome appears to be very similar, 98% of my reads mapped) and then calculate coverage either in BBmap itself or using pileup.sh

2) Assemble my reads using tadwrapper or tadpole and then map my reads to this assemblies and calculate coverage, again either in BBmap or pileup.

3) None of the above, I've gone wrong somewhere in my understanding.

Thanks for your help