I have 2 assemblies constructed from different libraries of same organism (PE reads 300bp and 150bp for Newbler and Platanus assembler, respectively). The Newbler assembly (followed by SSPACE and GapCloser) has better contiguity (60% higher N50, and much lower number of scaffolds) than Platanus assembly despite that the 300bp-library raw coverage fold is only 1/4 compared to 150bp-library used for Platanus. Could this be due to the difference between assembling approaches (OLC vs de Bruijn) ?
I tried to use GAM-NGS to merge 2 assemblies, but as I assigned Platanus assembly as Master and Newbler as Slave, the output out.gam.fasta is absolutely zero improvement compared to Master assembly, paired contigs built = 0 (as shown below)
[main] Loading contig sequences
master sequences loaded = 905200
slave sequences loaded = 17979
[merge] Writing contigs with no blocks to file: /work/gam-ngs_merge/out.noblocks.BF.fasta
[merge] Writing contigs with no blocks (after filtering) to file: /work/gam-ngs_merge/out.noblocks.AF.fasta
[merge] Paired contigs built = 0
Could someone give me some suggestion regarding this case? How can I exploit high-contiguity Newbler assembly to improve Platanus assembly?
Thank you very much in advance.