What is the difference between contigs and scaffolds in abyss? If contigs are put together (scaffolded?) from unitigs using PE reads, what's the difference between that and scaffolding?
Hi, Damian. The two stages are quite similar, but differ in their aims.
The contigging stage navigates the contig overlap graph to find paths between contigs linked by paired-end information, expands repeats and merges complete sequences without gaps. The contigging stage will only insert gaps where it believe there is a lack of sequencing data. It will not insert a gap to span a repeat that it cannot resolve. This algorithm is called SimpleGraph / MergePaths.
The scaffolding stage uses the contig overlap graph only incidentally and primarily uses the paired-end scaffold graph, which is the links between contigs inferred from paired-end reads. It does not expand repeats, and will insert gaps to scaffold over either sequencing gaps or repeats. This algorithm is called abyss-scaffold.
Cheers,
Shaun
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
I see. So the unitigging stage produces unambiguously merged k-mers. Then the contigging stage tries to resolve ambiguous cases in the k-mer graphs (junctions) via PE reads?
Yes, that's correct.