Kaether, K.-K.; Lemke, S.; Stadler, P. F.

Homology-based scaffolding is the task of ordering contigs based on their co-linearity with the DNA sequence of other, closely re- lated reference genomes. We investigate here how a recent approach to identify synteny anchors can be adapted to this task. The advantage of using anchors is that \"sufficiently unique\" anchor candidates can be precomputed independently for each genome assembly, including contig- level drafts. This allows an efficient identification of co-linear runs of anchor matches between a contig and a reference genome and, conse- quently, also order their contigs in relation to the reference. Such data can be improved by aggregating scaffolding results for multiple reference genomes. We present a protoypical pipeline based on the anchor-based synteny detection tool AncST and show that it produces scaffolds with accuracies comparable to competing tools at acceptable computational costs. Assuming that anchor candidates have been precomputed for the genomes of interest, the remaining computational efforts are on par with the fastest methods.