Genome Assembly
We performed a de novo hybrid genome assembly with MaSuRCA 3.3.1 (Zimin et al., 2017), using short (Illumina) and long (ONT) reads. Before the assembly step, we filtered the ONT reads with a Phred quality score (Q ≥ 5) using the NanoFilt software (included in NanoPack, De Coster, D’Hert, Schultz, Cruts, & Van Broeckhoven, 2018). Paired-end Illumina reads were parsed into MaSuRCA without any preprocessing, as adapters and errors are handled by the QuORUM error corrector (Marçais, Yorke, & Zimin, 2015), which is part of the MaSuRCA pipeline. MaSuRCA was run applying the following parameters: fragment mean (422), fragment stdev (312) and estimated genome size (1.2 Gbp). The resulting assembly was screened for contaminants with BlobTools v1.0 (Laetsch & Blaxter, 2017) -x bestsumorder. Assembly completeness was assessed with BUSCO 4.0.2 (Seppey, Manni, & Zdobnov, 2019) using the 8,338 single-copy conserved genes in aves_odb10 database (Kriventseva et al., 2019).