I've recently been trying different assemblers on my data. The latest one I tried is PERGA (http://www.ncbi.nlm.nih.gov/pmc/articles/PMC4252104/). It caught my attention because it uses machine learning to predict the correct extensions.
The authors report PERGA performing very well assembling S.pombe genome, which is what I'm doing. However, when I tried PERGA on my data (250 bp X 2.5m PE reads), I got poor results: scaffolds from PERGA had an N50 of ~30k bp while I was getting scaffolds of +110k bp N50 with other assemblers. The coverage was similar to that of of other assemblers. Has anyone had any experience with this tool?
The authors report PERGA performing very well assembling S.pombe genome, which is what I'm doing. However, when I tried PERGA on my data (250 bp X 2.5m PE reads), I got poor results: scaffolds from PERGA had an N50 of ~30k bp while I was getting scaffolds of +110k bp N50 with other assemblers. The coverage was similar to that of of other assemblers. Has anyone had any experience with this tool?