Hello,
I have run into this scenario one time too many:
Most (> 50 %) of genes are differentially expressed (FDR < 0.01) but the assumption for DESeq2 and edgeR is that most genes are not differentially expressed.
I have tried searching but can't find a proper discussion of this.
other small threads:
https://www.biostars.org/p/52183/
I'm doing bacterial RNA-seq with biological triplicates; 2 different data sets, with many different conditions and in both cases I have had > 50 % DE in most relevant pairwise comparisons.
Anyone have thought about how problematic violating the assumption would be or have alternative ideas for analysis?
Thanks.
I have run into this scenario one time too many:
Most (> 50 %) of genes are differentially expressed (FDR < 0.01) but the assumption for DESeq2 and edgeR is that most genes are not differentially expressed.
I have tried searching but can't find a proper discussion of this.
other small threads:
https://www.biostars.org/p/52183/
I'm doing bacterial RNA-seq with biological triplicates; 2 different data sets, with many different conditions and in both cases I have had > 50 % DE in most relevant pairwise comparisons.
Anyone have thought about how problematic violating the assumption would be or have alternative ideas for analysis?
Thanks.