Entering edit mode
3.7 years ago
yiren
▴
10
Hi everyone,
Now, I have one hundred rnaseq samples data, but this data is not strand specific. I did the differential gene expression and can I use this data to annotate gene structure of genome? or what can I do by the rnaseq? my data like below:
control1 case1(three replicates)
control2 case2(three replicates)
control3 case3(three replicates)
.
.
.
I have many cases and corresponding controls .How can i use these data do WGCNA?
Thank you very much
Thank you for your reply.I want to create a transcptional profile and find every condition differential expression gene which I will knock out in my strain.Now I have found the genes that are differential expression. So I want to know something else I can do by these RNAseq data?
Hi again, if you have already found your statistically significantly differentially expressed genes, then your goal is complete and you can proceed to design your knock-out experiment, no?
You can try WGCNA, in which case your input should be either the normalised counts or the transformed normalised counts (EdgeR's log2(CPM+1), or DESeq2's variance-stabilised or regularised log expression levels). What would be the goal of WGCNA, a network analysis workflow, in this situation?
Thank you very much.You are right and I am knocking the genes which I am interested in genes such as transcription factor. I try WGCNA because I want to make good use of the RNAseq data and I also want to display the transcriptional profile in network.I have normalized the counts by DESeq2's regularised log expression levels. My final normalised counts file like below:
I have twenty conditions about RNAseq and every condition has three biological replicates controls and three biological replicates case. Is it appropriate for my RNAseq data to do WGCNA? I have read some papers that they have one condition control(about 50 samples) vs case(about 50 samples).Thank you very much for your careful reply to me.
Hi again, can you share some of these other papers? Perhaps there would be issues using technical replicates, but are biological replicates also a problem for WGCNA? - I am not sure.
With RNA-seq, you can also do:
Hello again, The paper website is https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6788446/ .This is a clinical data paper.They did WGCNA used the data about normal samples(26 samples ) and case samples(70 samples).
I see... they have an additional problem of batch effects in that study (they use data from 3 separate studies). Best thing to do is proceed using all samples, and then try again using collapsed biological replicates. In my mind, however, it makes little sense to collapse biological replicates into a single sample.
Thank you kevin.I will try to to identify novel transcripts and gene signature creation as you said.I also try to do WGCNA using RNAseq data and I hope to discovery a transcriptional regulatory networks so that I can show my work that I sequenced many RNAseq data under many conditions in one figure.
Great - I look forward to seeing the results!