Repeat element annotation
We identified repetitive elements by integrating homology and de novo predictions. Homology-based transposable elements (TEs) annotation were obtained by interrogating a genome assembly with known repeats in the Repbase database v16.02 (Bao, Kojima, & Kohany, 2015) using RepeatMasker v4.0.5 (DNA-level) (Tarailo-Graovac & Chen, 2009) and RepeatProteinMask (protein-level; implemented in RepeatMasker). De novo TE predictions were obtained using RepeatModeler v1.1.0.4 (Smit & Hubley, 2010) and LTRharvest v1.5.8 (Ellinghaus, Kurtz, & Willhoeft, 2008) to generate database for a RepeatMasker run. Tandem Repeat Finder (v4.07) (Benson, 1999) was used to find tandem repeats (TRs) in the genome. A non-redundant repeat annotation set was obtained by combining the above data. Genome sequences of other marsupials were also interrogated using the repeat element annotation pipeline. The following were downloaded from DNA Zoo [Virginia opossum (Didelphis virginiana ; dv-2k (Olga Dudchenko et al., 2018))], the NCBI Reference Sequence Database (Release 86) (Pruitt, Tatusova, & Maglott, 2007) [gray short-tail opossum (Monodelphis domestica ; MonDom5; GCF_000002295.2), koala (Phascolarctos cinereus ; phaCin_unsw_v4.1; GCA_002099425.1), Tasmanian devil (Sarcophilus harrisii; Devil_ref v7.0, also known as sarHar1; GCA_000189315.1), Tammar wallaby (Notamacropus eugenii ; Meug_1.1; GCA_000004035.1), and common wombat (Vombatus ursinus ; GCF_900497805.2)], and the GigaScience Database [brown antechinus (Antechinus stuartii ; (Brandies et al., 2020a; Brandies, Tang, Johnson, Hogg, & Belov, 2020b)].