I am working with nematode and host Medicago. I have transcriptome data for both species and I also have a list of differentially expressed genes(DEG). Now I would like to see if DEG of host and nematode are associated by calculating a correlation between genes. Which package should I use? I am new in this area so please give me suggestions on network analysis, steps, packages, papers. Thank you!
Article clusterProfiler: an R Package for Comparing Biological Theme...
Dear Shova
I would say there are several approaches for this. For example, You can try a simple correlation analysis by computation of all pairwise correlations (many packages in R can do it, even built-in function "cor", which does have more limitations in comparison with, for example, "corr.test" or "rcorr") and then convert it to adjacent matrix for visualization. Or You can try WGCNA method (also in R), which can be more complicated for the beginner, but much more powerful for large data, which is perhaps Your case. Also, just to keep in mind, to be powerful, correlation requires enough observation points in Your dataset.
Dear Albert,
Thank you for your suggestion. Correlation requires enough observation points means number of replications?
I tried using corr.test in R but I got an error because my data arrangement is not correct. How should I arrange my data in excel?
Dear Shova
Yes, biological replicates are observation points. In Your specific case, I guess, You should have enough Nematode individuals and the same number of host individuals. And consequently the data has to be arranged in a way that will match specific nematode and its host pair.
Then, for R You just need to upload a data table in a form (if You want to receive a single correlation matrix with cross correlation of genes within and between species):
Name Gene_01 Gene_02 ... Gene_M
Nem_01
Nem_02
...
Nem_N
Host_01
Host_02
...
Host_N
Alternatively, You can separate Your data to two sets: one for nematodes, and one for host, and build two correlation matrices and then compare them. In this case, nematode-host matching is not needed. What approach to use depends on Your research question.
Thank you, Albert. I have transcriptome data for nematode and host for six-time points and DEGs I got are from all pairwise combinations between all time points.
Genes for plants and nematodes are also different. They do not have matching genes. My data looks complicated .
In this case, should I arrange my data like this:
T1 T2...........T6
Nema-gene1
Nema-gene2
.
.
.
Nema-gene-n
plant-gene1
.
.plant-gene-n
I appreciate your help.
- Badges
- Science topic
- Similar topics
- Papers