Hi Guys,
I have a RNA-seq data file for 10000 genes and read counts. the first column is geneID, 2nd one is read count for the control and the 3rd one is read counts for the treatment. I am going to compare two samples I have(one for control and one for the treatment). can i get P-values? if so, which statistical test can be performed? thanks
A good paper to refer to: https://genomebiology.biomedcentral.com/articles/10.1186/s13059-016-0881-8
Up-voted. Directly from the paper: "For small-scale studies that compare two samples with no or few replicates, the estimation of the negative binomial distribution can be noisy. In such cases, simpler methods based on the Poisson distribution, such as DEGseq [62], or on empirical distributions (NOISeq [10]) can be an alternative, although it should be strongly stressed that, in the absence of biological replication, no population inference can be made and hence any p value calculation is invalid."
In essence, yes you can get p-values but they'll be meaningless.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
Your results will be questionable, without replicates. Definitely need validation with independent technology.
Why do you have 10000 genes ? Are they standard annotations ? or picked up based on certain criteria ? It depends on how did you obtain the data .
i have got the read counts before using HTSeq
Which organism are you working with ? If its rna-seq data, probably you could try DESeq2 ( Read the section "without replicates") Or Gfold .