Subsampling option

[muffato]

Description of feature

The Snakemake pipeline has got an option for capping the coverage of the read data, subsampling if necessary. It may be useful to implement, especially for large genomes ?