Our current datastructure has a FP rate that depends on the kmer size. Its acceptable with k = 30 to 33, but is not adapted to smaller or longer k values. Thus we should also offer the possibility to use a classical Bloom filter for "non usual" k values
Our current datastructure has a FP rate that depends on the kmer size. Its acceptable with k = 30 to 33, but is not adapted to smaller or longer k values. Thus we should also offer the possibility to use a classical Bloom filter for "non usual" k values