Transcriptome and Mitochondrial Genome Assemblies
The assembled transcriptome was 297.65 Mb, encompassing 129,804 genes and 201,715 transcripts (Table 2) . Based on blast searches against full length reference proteins in the UniProt-SwissProt database, the transcriptome had 3,985 transcripts with >80% and 2,869 transcripts >90% alignment coverage on reference proteins. The BUSCO assessment of 2,442 single copy orthologs shows 97.4% completeness (complete and single copy =19.2%, complete and duplicated = 78.2%, fragmented = 1.6%, and missing = 1.0%). The high number of duplicated genes is due to the presence of several isoforms in the trinity output. When only the longest isoform per gene is used, the transcriptome size is 100.02 Mb (contig N50 = 3,991 bp, with an average and median contig length of 1,475 and 489 bp, respectively).
The mitochondrial genome of N. riversi is 15.8 Kb, with a full complement of 37 genes (Figure S1 and Table S2 ). This is substantially smaller than the published mitochondrial genome ofN. brevicollis (21.2 Kb), which has a large expansion of non-coding DNA near the replication origin. However, the mitochondrial genome of N. riversi is similar in size to the majority of Coleopteran mitochondrial genomes, and the gene order is largely conserved, except for a novel inversion with respect to other Carabidae that encompasses the nad2, trnI, trnM and trnW genes adjacent to the replication origin.