DINA Network Similarity

network similarity measures how closely two networks structurally resemble each other the dina network similarity analysis will compare the first network with all the other networks in mavatar discovery by quantifying the structural similarity based on node affinity and information flow using deltacon the analyses result in a similarity score and a p value between each pairwise comparison, visualized in a bubble chart with –log10(p value) on the x axis and the similarity score shown as variable bubble sizes when hovering over the bubbles, details of the comparison are given, as well as a warning if the overlap between the networks’ sample inclusion is large some datasets and samples are expected to overlap between different networks when two networks share a large proportion of their underlying data, a high similarity score may partly reflect that shared data rather than a true structural relationship between the networks, so pairs flagged with a high overlap warning should be interpreted with that in mind when your graph construction includes a second network, you can check the exact number of unique and shared samples between them on the dina network card a high similarity score will indicate functional similarities between the networks while if there are no significantly similar networks, it will indicate that the graph functions identified are highly specific to the network under investigation by right clicking on a network name, you can select add docid\ fnr1vl5zxsmm8jomxdyc5 either as first network or as second network, which could be useful for additional comparative pathway analyses read more about our network similarity calculations under dina network similarity analysis docid\ fx907vixmfpwg4dpj8 r5 and about our method validations under dina network similarity validations docid\ uhclpee8 fosx5lplvkqt