I'm currently analyzing RNA seq data and have two sets of conditions, control and knock out. I have triplicates of each, so 6 data entries per gene in total. I'm interested in identifying target genes. I have calculated fold changes and log2 fold changes, but I also need to find P-values to see if each is significant. I'm unsure what data to use to calculate these P-values? Should I use the raw data, fold changes, or log2 fold changes?
I'd suggest looking into tools designed for analysing RNAseq data. I generally use the R package DESeq2.
yes I second the answer by Greg Deakin
Or if u are not familiar with R u can use Galaxy server. (usegalaxy.org)
Well, you must be interested in finding the genes which are differentially expressed? This is a multiple hypothesis problem, therefore you have to consider the significantly expressed genes based on False Discovery Rate (FDR) cut-off and not p-value alone. This can be achieved by either edgeR or Deseq2 using R (if familiar) or usegalaxy.org RNAseq analysis pipeline.
You have to use the raw data if using any of the tools i.e., Deseq2 or edgeR.
This is one of the tutorials on the galaxy server using Deseq2:
https://training.galaxyproject.org/training-material/topics/transcriptomics/tutorials/ref-based/tutorial.html
Hope this helps.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python