Im doing an experiment where I'm trying to create a contig level assembly of a genome with soap denovo 2 by using a genome A. from running quast genome A had a very good value N50 i.e 1988282 contigs. but when I used the paired end genome A reads with soap denovo 2 with k51 to create a contig level assembly by running pregraph and contigs step the log file says that it has only 2641 contigs which very much lower than the original N50 value as seen in quast. how can I improve the contiguity with soap denovo2 ?