Fig. 1

Workflow scheme. First the reads are mapped to a reference sequence, called primary reference. Some of the mapping reads can be used to identify haplotype regions and link the reads corresponding to the odin and the source organellar sequence. Afterwards, a sequence is assembled for both the odin and the source organellar sequence for each region. Each of the assembled sequences is compared to a new reference sequence, called secondary reference, which may or may not be the same as the primary sequence. For each region, the sequences are assigned to be odins or source organellar based on the alignment score to the secondary reference. Finally, an organellar genome sequence is assembled, where the regions previously identified to contain odins now contain the chosen sequence from the alignment. The steps can be repeated in an iterative fashion using the resulting assembled sequence as the new primary reference sequence