Fig. 3

Fosmid pool assembly pipeline. For each fosmid pool, a single paired-end (PE) library sequenced at 2 x 150 bp was first filtered and trimmed of pNGS vector sequences, as well as those of Escherichia coli and other common contaminants, including Olea europaea chloroplast sequences. Reads were assembled with ABySS, gapfilled with GapFiller, and contaminants removed using a BLAST homology search. A consistency check was performed, breaking the assemblies at any point inconsistent with the proper insert size and orientation of fosmid pool PE reads. The resulting contigs were scaffolded using whole genome shotgun (WGS) data, followed by another round of gapfilling, decontamination and consistency checking, this time including the new WGS data. To repair the consistency broken assembly, a final round of scaffolding, gapfilling and decontamination was performed