Fig. 3: Host–virus genome read coverage ratios.

a Possible scenarios for the interpretation of genome read coverage results. If there is a low rate of viral replication, we expect the genome coverage ratio of the virus and host to be nearly the same in a metagenome. Higher viral coverage could result from a higher number of virions compared to host cells, or more copies of the viral genome in each infected cell, but could also mean that a lysogenic virus has more than the single predicted host species. Higher coverage of a bacterial genome suggests that single-cell genomics captured a very rare infection event or that the virus infected only a subset of cells (i.e. only certain strains). b Detection of the 59 de-replicated host–virus pairings in the three layers. The gray portions indicate the pairs in which virus, host, or both genomes were below the detection threshold. For 35 pairings (purple), detection was possible in at least one of the layers (>75% of the genome length covered). c Number of host genomes in each phylum for which the host–virus genome coverage ratio could (Detectable) or could not (Not detectable) be calculated. Virus and host icons indicate which one from the host–virus pair was above the detection threshold. d Fold-change of host and virus genome coverage for the four pairs detected in two or three layers. The dashed line indicates the 1.5× fold-change range. Dots positioned on the right from the host-baseline in the middle indicate higher coverage of the viral genome, while on the left indicate lower coverage of the viral genome compared to the host. e Fold-change of host and virus genome coverage of all 35 host-virus pairs (dots) grouped by host phyla. The distribution of points relative to the x-axis is described in (d).