Fig 1.
Sankey diagram summarizing the fungal host and virus state of the SRA samples queried in this study and recognized mycoviral sequences.
The height of the bars in each column is indicative of the number of samples. Classification of fungal host is summarized in the two, leftmost columns. Column “Fungal Class”: fungal species are sorted into one of eight taxonomic classes, all genera from the same class are the same color; Column “Fungal Genus”: separates the fungi into their respective genera; colors used in this column are to highlight the individual genera but similar colors have no relation to one another. The third column, “Data Source” is color-coded by the origin of the sample; blue indicates a known mycovirus, and the colors pink and grey indicate an SRA sample was analyzed where either a virus sequence was detected (pink) or was not detected (grey). The two rightmost columns summarize the information related to virus taxonomy. Column 4: “Viral Family” organizes the samples that have either a known or discovered virus by viral family. Column 5: “Viral Class” summarizes the viral genomes by nucleic acid and/or strandedness. SRA sequencing samples without a virus detected are in the category “No Virus”.
Table 1.
Summary of mycoviral genomes identified from fungal RNA-seq datasets.
Fig 2.
Phylogram of the RdRP-containing ORF for viruses classified as dsRNA viruses.
A maximum likelihood phylogenetic tree was generated with RaxML, using the LG+G+I amino acid substitution model. Scale bar represents 1.0 amino acid substitutions per site and numbers at the nodes indicate bootstrap support over 70% (1000 replicates). The sequence from Simian Rotavirus A (SRVA) serves as the outgroup. Three officially recognized families are present: Partitiviridae, Chrysoviridae, and Totiviridae and are highlighted with a green, orange, or blue box respectively. Gray boxes surround families currently unrecognized. The structure of the genomic segment(s) for each highlighted group is depicted below the family name. Genera within the Partitiviridae, Chrysoviridae and Totiviridae family are indicated. Viruses identified in this study are noted with a red star. RdRP: RNA-dependent RNA Polymerase; CP: Coat Protein; HP: Hypothetical Protein.
Fig 3.
Analyses of identified novel non-segmented RNA viral genomes in relation to Amalgaviridae and Partitiviridae.
(A) Phylogram of the ORF containing the RdRP motif for members of the new mycovirus group along with select partitiviruses, totiviruses and amalgaviruses; sequence from Penicillium chrysogenum virus (PcCV) serves as the outgroup. Full names and accession numbers for protein sequences used in the alignment are in S2 Table. Scale bar on the phylogenetic tree represents 1.0 amino acid substitutions per site and numbers at the nodes indicate bootstrap support over 70% (1000 replicates). Diagrams illustrating the genome structure and predicted ORFs are present for each group of viruses. (B) Percent identity matrix, generated by Clustal-Omega 2.1, of new mycoviral family, partitiviruses, and amalgaviruses. The top half of the matrix, in the blue scale, is the percent identity of the RDRP protein coding sequence and the bottom half of the matrix, in the green scale, is the percent identity of the putative CP. For sake of clarity, the 100% identity along the diagonal has been removed and outlined stars are used to note junction point for the row and column of the viruses identified. Only RdRP values ≥50% and CP values ≥30% are indicated. (C) Density plot of read counts per nucleotide across the viral genome for BbNRV1. Graphical depiction of the genome structure, ORFs and predicted motifs are on along the top and a heatmap of normalized read counts is on the bottom. Read counts are normalized to a 0 – 1 scale, where the maximum read count for a given genome equals 1. Total reads mapped to the virus genome and average read coverage are noted. The red star indicates the virus identified in this study. RdRP: RNA-dependent RNA Polymerase; CP: Coat Protein; ORF: Open Reading Frame.
Fig 4.
Analyses of Aspergillus heteromorphus alternavirus 1 (AheAV1), a member of the recently proposed family Alternaviridae.
(A) Percent identity matrix, generated by Clustal-Omega 2.1, of alternaviruses and members of the Chrysoviridae family. The top half of the matrix, in the blue scale, is the percent identity of the RdRP protein coding sequence (from the RNA1 segment) and the bottom half of the matrix, in the green scale, is the percent identity of the second largest genomic fragment, encoding GP1. For sake of clarity, the 100% identity along the diagonal has been removed and outlined stars are used to note junction point for the row and column of the virus identified. Values ≥20% identity are noted. (B) Density plot of read counts per nucleotide across the viral genomic segments of AheAV1. Graphical depiction of the genome structure, predicted ORFs and motifs for each segment are on top and a heatmap of normalized read counts is below each diagram. Read counts are normalized to a 0 – 1 scale, where the maximum read count for a given genome equals 1. Total reads mapped to the virus genome and average read coverage are noted. (C) Alignment of 5’UTR sequences of the three genomic segments of AheAV1. Conserved nucleotides are highlighted in blue. (D) Alignment of RdRP domains of Alternaviruses found in (A). The eight conserved RdRP motifs of dsRNA polymerases are noted along the top of the alignment and the ADD triad, found only in Alternavirus sequences, is highlighted in green. The number of amino acids not shown in alignment are noted in parentheses. Viruses identified in this study are indicated with a red star. RdRP: RNA-dependent RNA Polymerase; GP1: Gene Product 1.
Fig 5.
Analyses of identified viral genomes belonging to the two proposed genera, Epsilonpartitivirus and Zetapartitivirus, within the Partitiviridae family.
(A) Percent identity matrix, generated by Clustal-Omega 2.1, between members of the two new groups and gammapartitiviruses, the most closely related, ICTV-recognized genus. The top half of the matrix, in the blue scale, is the percent identity of RdRP protein sequences and the bottom half of the matrix, in the green scale, is the percent identity of CP protein sequences. For sake of clarity, the 100% identity along the diagonal has been removed and outlined stars are used to note junction point for the row and column of the viruses identified. Two cutoff values were applied and as such only percent identities ≥29% among RdRP sequences and ≥17% among CP sequences are indicated. Virus names in italics refer to viruses isolated from insect, not fungal, hosts. Bolded virus names are member species recognized by ICTV. (B and E) Density plot of read counts per nucleotide across the viral genome segments for the two epsilonparitiviruses (B) and two zetapartitiviruses (E). Graphical depiction of the genome structure, predicted ORFs, and RdRP motif are along the top and a heatmap of normalized read counts is below. Read counts are normalized to a 0 – 1 scale, where the maximum read count for a given genome equals 1; heatmap scale found in (E) applies also to (B). Total reads mapped to the virus genome and average read coverage are noted for each virus. (C) Alignment of the 5’ UTR sequences from the two genomic segments of the new species of epsilonpartitivirus, CePV1. Conserved nucleotides are highlighted in blue and numbers in parentheses indicate nucleotides not shown. (D) Alignment of the six conserved motifs within the RdRP protein sequence of all four proposed zetapartitiviruses. Amino acids highlighted in blue are only common among zetapartitiviruses while residues in grey are shared with gammapartitiviruses. (F and G) The 5’UTR (F) and 3’UTR (G) sequences from both genomic segments for all four zetapartitiviruses were aligned using MAFFT and a graphical representation of the most highly conserved sequences found is depicted as a sequence logo, generated using WebLogo version 3, whereby the height of a nucleotide indicates the sequence conservation at that position (http://weblogo.threeplusone.com/). A consensus sequence for each UTR is indicated below in grey. Red stars indicate viruses identified in this study. PV: partitivirus; RdRP: RNA-dependent RNA Polymerase; CP: Coat Protein.
Fig 6.
Phylograms for three different families of (+)ssRNA viruses.
RdRP motifs from the Endornaviridae family (A), Narnaviridae and Ourmiaviruses (B), and Hypoviridae (C) family were used to construct maximum likelihood phylogenetic trees generated with RaxML, using the LG+G+I amino acid substitution model. Full names of viruses and sequences can be found in S2 Table, including the outgroup, Potato Virus Y (PVY). Bolded names are ICTV-recognized member species while italicized names are related, but unclassified viruses. Full names and accession numbers for protein sequences used in the alignment are in S2 Table. Scale bar on the phylogenetic tree represents 1.0 amino acid substitutions per site and numbers at the nodes indicate bootstrap support over 70% (1000 replicates). Viruses identified in this study are noted with a red star.
Fig 7.
Analyses of Acidomyces richmondensis tobamo-like virus 1 (ArTlV1) and related viruses.
(A) Phylogram of the RdRP-motif for ArTlV1, related mycoviruses, and select members of the genus Tobamovirus and family Bromoviridae; sequence from Rubella Virus (RV) serves as the outgroup. Bolded names are ICTV-recognized member species. Full names and accession numbers for protein sequences used are in S2 Table. Numbers at the nodes indicate bootstrap support values ≥65% (1000 replicates). Scale bar on the phylogenetic tree represents 1.0 amino acid substitutions per site. Diagrammatic genome structures are to the right of the phylogram. (B) Density plot of read counts per nucleotide across the ArTlV1 viral genome. Graphical depiction of the genome structure, predicted ORFs, and encoded motifs are along the top and a heatmap of normalized read counts is on the bottom. Reads are normalized to a 0–1 scale, where the maximum read count equals 1. Total reads mapped to the virus genome and average read coverage are noted. (C) Percent identity matrix, generated by Clustal-Omega 2.1. The top half of the matrix, in the blue scale, is the percent identity of the RdRP motif sequence while and the bottom half of the matrix, in the green scale, is the percent identity of the residues from Viral Helicase motif. For sake of clarity, the 100% identity along the diagonal has been removed and outlined stars are used to note junction point for the row and column of ArTlV1. Cutoff values of 44% and 40% were applied to the RdRP and Helicase matrices respectively. Viruses identified in this study are noted with a red star. VMt: Viral Methyltransferase; VHel: Viral Helicase; RdRP: RNA-dependent RNA Polymerase; DH: DEXDc Helicase; ORF: Open Reading Frame.
Fig 8.
Analyses of four newly identified (+)ssRNA viral genomes and viruses belonging to the proposed family Ambiguiviridae.
(A) Phylogram of the RdRP-motif for four new viruses and related viruses along with select members of the family Tombusviridae and members of an unnamed group of (+)ssRNA insect viruses; sequence from Tobacco Mosaic Virus (TMV) serves as the outgroup. Bolded names are ICTV-recognized member species while italicized names are recognized, but unclassified viruses. Full names and accession numbers for protein sequences used are in S2 Table. Numbers at the nodes indicate bootstrap support values ≥60% (1000 replicates). Scale bar on the phylogenetic tree represents 1.0 amino acid substitutions per site. Diagrammatic genome structures and predicted ORFs for each clade of viruses is to the right of the phylogram. (B) Density plot of read counts per nucleotide across four viral genomes. Graphical depiction of the genome structure, predicted ORFs, encoded motifs and the amber stop codon (TAG) for each genome are along the top and a heatmap of normalized read counts is on the bottom. Reads are normalized to a 0–1 scale, where the maximum read count for a given genome equals 1. Total reads mapped to the virus genome and average read coverage are noted. (C) Multiple sequence alignment of the RdRP motifs of members of the Ambiguiviridae family. Motifs A-E found in (+)ssRNA viral RdRP sequences are indicated along the top of the alignment. Conserved residues that differ from canonical (+)ssRNA RdRP sequences, including the unusual GDN triad found in Motif C, are highlighted in blue. Numbers in parentheses are residues not shown. (D) Percent identity matrix, generated by Clustal-Omega 2.1, of mycoviruses that comprise the proposed family Ambiguiviridae. The top half of the matrix, in the blue scale, is the percent identity of the RdRP motif sequence and the bottom half of the matrix, in the green scale, is the percent identity of the ORF1 amino acid sequence. For sake of clarity, the 100% identity along the diagonal has been removed and outlined stars are used to note junction point for the row and column of the viruses identified. All RdRP motif sequences share ≥35% identity and ORF1 protein sequences share ≥15% identity. Viruses identified in this study are noted with a red star. RdRP: RNA-dependent RNA Polymerase; ORF: Open Reading Frame; HP: Hypothetical Protein.
Fig 9.
Analysis of Aspergillus homomorphus yadokarivirus 1 (AhoYV1) and Penicillium digitatum yadokarivirus 1 (PdYV1) viral genome sequences, members of the proposed family Yadokariviridae.
(A) A phylogram of the RdRP coding sequence from members of the Yadokariviridae family and select (+)ssRNA viruses from the families Caliciviridae, Fusariviridae, and Hypoviridae; sequence from Potato Virus Y (PVY) serves as the outgroup. Virus names have been abbreviated; full names and accession numbers for protein sequences used in the alignment are in S2 Table. Phylogenetic tree was generated with RAxML, using the LG+G+I amino acid substitution model. Scale bar represents 1.0 amino acid substitutions per site and numbers at the nodes indicate bootstrap support over 70% (1000 replicates). To the right of the phylogram are representative diagrams of the genome structures for each family. (B and C) Density plot of read counts per nucleotide across the viral genome for (B) AhoYV1 and (C) PdYV1. Graphical depiction of the genome structure, the predicted single ORF and the RdRP motif are along the top and a heatmap of normalized read counts is on the bottom. Read counts are normalized to a 0–1 scale, where the maximum read count for a given genome equals 1. Legend for (C) is the same as the legend found in (B). Total reads mapped to the virus genome and average read coverage are noted for each virus. (D) Percent identity matrix generated by Clustal-Omega 2.1 using the RdRP motif sequences from the viruses analyzed in (A). For the sake of clarity, only the top half of the matrix is included and the 100% values along the diagonal have been removed. Percent identities greater than 25% are indicated. Viruses identified in this study are noted with a red star. RdRP: RNA-dependent RNA Polymerase; ORF: Open Reading Frame.
Fig 10.
Bioinformatics steps of the virus-like RdRP discovery pipeline.
Computational program names are highlighted in blue. Reads used in the bowtie2 alignment step were either as-is or adaptor-trimmed by cutadapt; details in the Materials and Methods. The dashed line indicates the workflow for fungi where no genome sequence was available; instead of running bowtie2, the reads were fed directly into a Trinity de novo assembly.