Assuming RAMLIMT is not exceeded, how do you calculate the expected memory usage of hmmsearch given a sequence database and a hmm file? The manual says sequence length * database length. Do we just substitute model length in the above equation for hmmsearch memory footprint? Ta, Dan.