Hi all,
I wonder if there is a way to extract the edges between intra-group nodes in relevance network? For example, I have 2 omics data, one is based on amplicon sequencing of bacteria 16S rRNA and one is based on microeukaryotic 18S rRNA. I want to get not only bacteria-microeukaryotic associations but also bacteria-bacteria and microeukaryotic-microeukaryotic. Actually, in the real dataset, I have up to 4 groups, and I use diablo approach, so I want to extract 4 intra group associations as well.
hi @adyari808,
Unfortunately in our integrative approach, we are mostly interested in the inter associations.
One potential trick would be
- extract the name of the variables selected (selectVar()), then calculate the correlation within each omics (perhaps put a high threshold)
- extract the M matrix from your network result (sim.mat = network(nutri.res); sim.mat $M)
- combine those matrices (I dont know how) and input into cytoscape.
Or maybe cytoscape already has a function to calculate the correlations inter and intra based on step 1 above? Basically the idea is to use the mixOmics method you used as a filtering step for further downstream analysis.
Kim-Anh