FIGURE 2.

Genome organizations and phylogenetic analyses of wild bee viruses and insect-specific viruses distantly related to plant viruses. (A) The genome structure with in silico predicted ORF annotations (gray arrow boxes) on the assembled viral transcripts (black lines). In case a genome was not fully recovered by RNA-Seq, the probable complete genome sequence is represented by a gray dashed line. The sequence length in nucleotides is given by the numbers below the lines. Protein domains were predicted using interproscan and are depicted by yellow boxes on top of the ORFs. In blue is the coverage by re-alignment of RNA-Seq reads to the contigs. Below left each map is the sequencing library name wherein the virus was detected and it is underlined in case the depicted transcript originated from that library. ^∗In case of O. cornuta nege-like virus, the contig was obtained by the combined de novo assembly of all Ocor datasets. Below right is the normalized coverage corresponding to the heatmap depicted in Figure 1A. (B) Maximum likelihood trees based on the partial replicase protein (RNA viruses) or non-structural protein 1 (densoviruses). Red abbreviations correspond to the new virus(es) for which a genome organization is given in (A). Blue abbreviations correspond to viruses also detected in the current study but not illustrated in (A). Green abbreviations correspond to known bee pathogenic viruses (not detected in this study). Virus full names and protein accession numbers are included in Supplementary File S2.