Insights into the global freshwater virome

Abstract

Viruses are by far the most abundant life forms on this planet. Yet, the full viral diversity remains mostly unknown, especially in environments like freshwater. Therefore, we aimed to study freshwater viruses in a global context. To this end, we downloaded 380 publicly available viral metagenomes (>1 TB). More than 60% of these metagenomes were discarded based on their levels of cellular contamination assessed by ribosomal DNA content. For the remaining metagenomes, assembled contigs were decontaminated using two consecutive steps, eventually yielding 273,365 viral contigs longer than 1,000 bp. Long enough contigs (≥ 10 kb) were clustered to identify novel genomes/genome fragments. We could recover 549 complete circular and high-quality draft genomes, out of which 10 were recognized as being novel. Functional annotation of these genomes showed that most of the annotated coding sequences are DNA metabolic genes or phage structural genes. On the other hand, taxonomic analysis of viral contigs showed that most of the assigned contigs belonged to the order Caudovirales, particularly the families of Siphoviridae, Myoviridae, and Podoviridae. The recovered viral contigs contained several auxiliary metabolic genes belonging to several metabolic pathways, especially carbohydrate and amino acid metabolism in addition to photosynthesis as well as hydrocarbon degradation and antibiotic resistance. Overall, we present here a set of prudently chosen viral contigs, which should not only help better understanding of freshwater viruses but also be a valuable resource for future virome studies.

Keywords: auxiliary metabolic genes; bacteriophages; freshwater; metagenome; virome.

Figure 1

Distribution of contigs according to their clustering result based on their shared proteins using vConTACT2.

Figure 2

Intersection between different biomes based on common contigs with nonzero abundance. The figure was generated using upset function of UpSetR package v.1.4.0 (nsets = 8, order.by = “freq,” nintersects = 50, all other arguments set to default; Conway et al., 2017).

Figure 3

Heatmap of the abundance of 29 contigs with nonzero abundance among all studied biomes. The heatmap was generated using Heatmap3 R package v.1.1.7 (Zhao et al., 2014) in R v.4.0.2 (R Development Core Team, 2020) after log₁₀ transformation of abundance values for better visualization.

Figure 4

Phylogenomic tree of the 29 contigs with nonzero abundance among all studied biomes. The numbers above branches are GBDP pseudo-bootstrap support values from 100 replications. The branch lengths of the resulting VICTOR tree are scaled in terms of the D4 distance formula used.

Figure 5

Gene map of freshwater_SRR107147.contig000000001. CDS highlighted in cyan are hypothetical proteins, while the purple ones have labeled annotations.

Figure 6

PsbA phylogenetic tree. Red circles at the tip of branches denote sequences identified in the current study, while blue circles denote RefSeq PsbA sequences collected from NCBI. Two main clades can be observed, whose branches are colored in green and black.