Fig. 1

Sequence-to-graph alignment with SPAligner. Alignment of query sequence S (orange bar) to assembly graph G. Assembly graph edges are considered directed left-to-right (explicit edge orientation was omitted to improve clarity). Upper-left: Anchor search. Anchor alignments (regions of high similarity) between the query and the edge labels are identified with BWA-MEM. Upper-right: Anchor filtering. Anchors shorter than K (anchors 2, 6, 11), anchors “in the middle” of long edge (anchor 7) or ambiguous anchors (anchor 10 mostly covered by anchor 9, both anchors 4 and 5) are discarded. Bottom-left: Anchor chaining. Heaviest chain of compatible anchors (chain 3→8→9) is determined. Bottom-right: Reconstructing the filling paths. Paths for fragments of the query between the consecutive chain anchors (as well as left- and right-most fragments) are reconstructed