Miseq reads

Raw DNA sequence reads from all sequencing platforms were filtered and trimmed to ensure quality and then depleted of adapter sequences, and paired-end transcriptomic reads were merged using the CLC Genomics Workbench v.6.5 (see File S2 in the supplemental material). The mitochondrial genomes of both Perkinsela strains were assembled from combined next-generation sequencing reads with the Newbler Assembler (GS de novo Assembler v.2.9), from single 454, mate-pair, and paired-end Illumina HiSeq reads in the case of strain CCAP1560/4 and from paired-end Illumina MiSeq reads for strain GillNOR1/I (see File S2). A number of assembly parameters were tested with the goal of maximizing mitochondrial contig size. Manual analysis of a graph of alternative contig connections (produced by Newbler) with an in-house visualizing script was used to close gaps and assemble long repetitive regions. RNA-seq assemblies were performed with Trinity software (75 (link)).

PacBio reads for Foc isolate Fus2 were assembled using Canu and polished using Illumina MiSeq reads in Pilon to correct erroneous SNPs and InDels^{37 ,38 (link)}. De novo assembly of MiSeq data for the remaining six genomes was performed using Spades v.3.5.0^{39 (link)}. In all cases Quast^{40 (link)} was used to summarise assembly statistics and BUSCO^{41 (link)} used to assess completeness of gene space within the assembly. Assemblies were edited in accordance with results from the NCBI contamination screen (run as part of submission to Genbank in November 2016) with contigs split, trimmed or excluded as required. RepeatModeler, RepeatMasker and transposonPSI were used to identify repetitive and low complexity regions (http://www.repeatmasker.org, http://transposonpsi.sourceforge.net). In addition to generating de novo assemblies, Illumina sequencing reads were mapped to the PacBio Fus2 assembly. Alignment was performed using Bowtie2 v.2.2.4 before bedtools-intersect was used to determine number of reads aligning across 100 Kb windows in Fus2 contigs^{42 (link),43 (link)}.

Genomic DNA of 30 C. perfringens strains was extracted (Keto-Timonen et al., 2006 (link)), and whole-genome sequencing was performed using PacBio RSII (Institute of Biotechnology, Helsinki, Finland). Sequenced genomes were assembled using HGAP3 and checked for circularity using Gap4 (Staden et al., 2003 (link); Chin et al., 2013 (link)). To improve the draft assembly, Illumina MiSeq reads and Pilon tool were used for genome polishing (Walker et al., 2014 (link)). Sequenced C. perfringens genomes were deposited in the GenBank (accession numbers listed in Supplementary Table 3). Both sequenced and downloaded genomes were annotated using Prokka (Seemann, 2014 (link)). Protein functional annotation and conserved domain predictions were refined using CD-search and CDD (Li and Godzik, 2006 (link); Marchler-Bauer et al., 2009 (link)).