Aligner 6

Raw data were filtered to remove sequences with low quality scores (Q<20) and exact duplicate reads. The ends of the reads with low quality were trimmed using PRINSEQ [14 (link)]. The host genome was removed by mapping to the Anopheles, Aedes and Culex genomes with Bowtie2 [15 (link)]. Unmapped reads were extracted and then subjected to a basic local alignment search tool (BLASTn and BLASTx) querying the NCBI nucleotide (nt) and protein sequence database (nr) using an E-value cut-off of 1e-03. The identified viral reads were assembled by de novo assemblers using CodonCode Aligner 6.0.2 (CodonCode Corporation) and SeqMan NGen 11.2.1 (DNASTAR) to generate longer sequences. Alignment of contigs and unassembled reads to viral reference genomes was performed using CodonCode Aligner 6.0.2 (CodonCode Corporation).

Due to DNA degradation in the very old specimens of A. joubini and the British A. crataegi, the 658 bp COI barcode region was recovered using an NGS-based protocol^{27 (link)}. In brief, for each sample, multiple short, overlapping amplicons were generated using nested, multiplex PCR. In order to associate reads with their source specimen, the amplicons were tailed with sample-specific universal molecular identifiers (UMI) before being pooled for sequencing on an Ion Torrent PGM. The short sequence reads were attributed to a sample via the UMIs, and filtered for quality and length before being assembled into a single barcode sequence.
For the remaining (younger) specimens, PCR and Sanger sequencing were carried out following standard procedures for Lepidoptera^{28 (link),29} . Briefly, the 658 bp barcode region was amplified using the primers LepF1 (5′ ATTCAACCAATCATAAAGATATTGG 3′) and LepR1 (5′ TAAACTTCTGGATGTCCAAAAAATCA 3′) and sequenced on an ABI 3730XL (Applied Biosystems capillary sequencer). The trace files were edited using CodonCode Aligner 6.0.2 (CodonCode Corporation, Dedham, Massachusetts) and all resulting mtDNA sequences were aligned using the same program. All sequences were submitted to GenBank (Accession Numbers in Supplementary Dataset File) and BOLD system repository (dataset name: DS—APORIA; https://doi.org/10.5883/DS-APORIA).

The genomic sequences of ZmbHLH16, including its 5′ and 3′ untranslated regions (UTRs), were amplified from 78 maize inbred lines (see Table S4 for details) with the primers 5′-GGAAGGAGGAAACCAAGTCG-3′ and 5′-TGTAACGAGCAAGCGGATTTA-3′. PCR was performed according to the manufacturer's protocol using high-fidelity polymerase KOD FX (Toyobo). PCR-amplifying fragments were purified and sequenced directly using an ABI 3730XL DNA Analyzer manufactured by Tsingke Biotech. After ambiguous sequences were manually deleted, the sequence polymorphisms of ZmbHLH16 among the 78 maize inbred lines were analyzed using CodonCode Aligner 6.0.2 software (CodonCode Corporation, Dedham, MA, USA). For molecular evolution analysis, certain parameters were calculated as follows: (1) the nucleotide diversity of common pairwise nucleotide difference per site (π) with DnaSP 5.0 (Librado and Rozas, 2009 (link)); (2) in neutrality tests, the evolutionary pressure in ZmbHLH16 via Tajima's D test (Tajima, 1989 (link)) and Fu and Li's statistics (Fu and Li, 1993 (link)); (3) the LD matrix of ZmbHLH16 was characterized by evaluating r² values based on SNPs and InDels (MAF≥0.05) in TASSEL 2.0 (Bradbury et al., 2007 (link)). An LD plot was obtained in Haploview 4.2 (Barrett et al., 2005 (link)), and the LD decay was assessed by averaging r² values with a distance of 250 bp.