Have little experience with metagenomes vs genomes, so this post is probably full of stupid questions. My thinking is to do a first pass assembly, map reads to the scaffolds and directly assess the discontinuities between scaffolds. If necessary, was thinking of mapping PE reads from other samples to evaluate the linkages between contigs, in the event that contigs weren't scaffolded due to insufficient coverage.
The other alternative is to combine some PE runs, which gives me more pairs and perhaps more coverage on the graph, but introduces the possibility of more graph complexity than is required. I'm not a fan of the latter approach, but I am curious as to the "best" way to tackle this problem.
The third is to finish reading Mad Albertsen's paper on differential coverage binning...any other suggested reading would be greatly appreciated.
The other alternative is to combine some PE runs, which gives me more pairs and perhaps more coverage on the graph, but introduces the possibility of more graph complexity than is required. I'm not a fan of the latter approach, but I am curious as to the "best" way to tackle this problem.
The third is to finish reading Mad Albertsen's paper on differential coverage binning...any other suggested reading would be greatly appreciated.