In the same VC but without Interactions

[Former user]

I guess if two virus are in the same VC, it must means that they have some protein clusters in the same and that they must have the interaction. But in my vContact2 output file I found that two viral contigs with two database virus are in the same VC but they don't have the interactions. That's quite confusing, could you please explain why or is that I have some misunderstanding about the result?

Attachments are the genome_by_genome_overview.csv file result, is that because of the low Topology Confidence Score?

Thanks a lot !!

[Ben Bolduc]

Yes, viruses in the same VC must share a minimum number of protein clusters (at least 3, but usually many more). In vConTACT2, when genomes share many protein clusters, we’re assuming that the more PCs shared means the more phylogenetically related they are. For example, if Genome A and Genome B share 5 of their 50 PCs, they might be related. However, if Genome A shares 45 of their 50 PCs with Genome C, then they’re highly likely to be related. And to go one step further, Genome A and Genome C are probably in the same genus, but Genome B might only be in the same order or family.

In your example, virus1 and virus2 are likely to be in the same genus. Since the quality is low (as is the topology confidence score, which takes into account quality), it may indicate there are other genomes (that aren’t in VC_159) connected to virus1 and virus2.

When you mention that they have “no interactions” - are you referring to connections within the network, or some biological activity (like a helper virus, or superinfection)? The only inference vConTACT2 can make is that any two genomes being likely taxonomically related. Definitive proof would require more intensive bioinformatic approaches.

-Ben

[li dafeng]

Thanks a lot Ben:

Yes, the interactions means the connections within the network. I found in my Cytoscape network result there is no connections between the virus 1 and the Bacteria phage APSE1 and 2 in the VC_159 and the virus 1 would have the connections with other viral genomes. Meanwhile, I used the tblastn to investigate the protein similarity between the virus1 and Bacteria phage APSE1, and I found there is only some matches between the protein share by the viral genomes (some fragments of the protein, not the whole protein), so as u said there are other genomes (that aren’t in VC_159) connected to virus1 and virus2 and I do find some connections between the virus 1 and other genomes in the network.

So, I arise the question why they would be clustered together but share less proteins (no connections within the network and tblastn with low similarity). May be as you said vcontact2 is considered the radio of the shared protein not only the amount?