I have six samples treated under 0h, 8h, 24h, 48h, 72h, 96h. I have the fpkm value of genes calculated by tuxedo. The data was reshaped as each row corresponds to gene and column to sample.
The data was handled by WGCNA according the steps of tutorial on the internet. I have got same modules from the wgcna results. In one module, I find out the fpkm value according the genes in this module. But the fpkm value of same genes are weird. The data below is part of the values:
Csa7G200810 Csa1G073690 Csa2G005330 Csa3G177990 Csa6G518250 Csa4G496230
PM.0 0.000000 0.0000 0.000000 0.0000 0.000 0.0000
PM.8 0.736315 0.0000 0.000000 0.0000 0.000 0.0000
PM.24 0.000000 0.0000 0.000000 0.0000 0.000 0.0000
PM.48 0.114667 0.0000 0.513395 0.0000 0.000 31.6024
PM.72 1.271810 17.9989 0.921125 13.1389 145.797 42.5157
PM.96 1.276460 14.3995 1.083880 15.4605 167.235 48.9445
the value of Csa6G518250 Csa4G496230 are different from the others at PM.72 and PM.96. the two genes do not look like co-express according the fpkm value, the fpkm value at PM.72 and PM.96 are greater then PM.0~ PM.48. I think co-expressed genes' fpkm value should be in a small scope.
How can I explain the fpkm value of the two genes?
What makes you say that those don't appear to be coexpressed? They all follow a similar profile in terms of expression time-course and have greater magnitude than the others there. Perhaps one could argue that Csa1G073690 should be in there as well (have a look at the module membership values of these, I expect it'll be informative).
BTW, "I think co-expressed genes' fpkm value should be in a small scope" doesn't really parse. I suspect that you mean something else by "scope", though I'm not sure what it is from context. Perhaps that you would expect genes said to covary to have more similar FPKMs?
Sorry about the confusing, I think the fpkm values of co-expressed genes should be similar. I found i have made a big mistake, these samples were treated as biological replications instead of time couse. Thank you for you advise Devon.