Should I Be Using K-Mer Coverage In Velvet Parameters Instead Of Nucleotide Coverage?

Question

[Panos]

Is Velvet using k-mer coverage values instead of nucleotide coverage in its parameters?

Suppose, for example, that my average read length is 36bp and k-mer size is 21. If I want to eliminate all regions with coverage of 2x or below should I use '-covcutoff 2' or '-covcutoff 0.9' (the latter based on a formula in the Velvet manual which converts nucleotide coverage to k-mer coverage - see Section 5.1 in the manual)? What about the '-exp_cov' parameter? Should I use k-mer coverage values for this parameter, too?

Answer 1

[Jeremy Leipzig]

yes 0.9, everything is kmer-coverage (read kmers per contig kmers), my understanding is that nothing is Velvet is x-coverage based, including exp-cov