Hi there,
I am doing smallRNA data analysis and I have a matrix of counts 17 miRNA and 16 samples I wonder if its correct to perform the results() from DESeq2 to get the DEM's.
Is there a minimum number of miRNA that is correct to perform the analysis with DESEq2 package?
Please, can anyone help me about this doubt?
Thanks in advance
Andreia
With 17 features, I would first doubt the accuracy of the size factors. The dispersion estimates will be decent because of the 16 samples, but if you don't know what is "housekeeping" then there's no hope for DE analysis with so few features.
I tested TMM and quantile normalization like some papers about smallRNA tested, actually i think there is no THE BEST normalization method :/ for this type of the analysis, the same thing with the aligner (i tested bwa and bowtie for this dataset with similar results).
In my particular case its harder, the data that came out its not helping i'm afraid.
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
What kind of assay is this? With only 17 miRNAs it is probably not NGS, is it? What are the values, are these counts?
Yes they are raw counts.
I know that is very small data! I did in past 2 data smallRNAseq analysis (not alot experience but i have a steady workflow...) but nothing with this quality. I already told the client about this.
This is from semen samples and the wetlab protocol is hard to get a good quality and quatity of RNA... :/
And yes is NGS :)