Dear all,

Our experiments is based on gene expression of a single tissue with contrasts between 3 treatments of 3 replicates. Each set of replicates was sampled from the same population and prepared and sequenced concurrently.

We have performed RNASeq as follows:

1. Align reads to reference with Salmon
2. Import data to DESeq2 with tximport
3. Identify differential expression with a (varying) log-fold change using apeglm, resulting in S-values and a significance threshold of 0.005 was applied.

We have repeated this experiment in three species using the same R code, and in one of the species the MA plots indicate a tiny amount of fold-changes for the majority of genes in the experiment; this is true for all three possible contrasts for that species. MA plots look normal for the other species. We are trying to identify or rule out a technical explanation for this result as we have not seen similar plots before. Is anyone able to offer their opinions or hypotheses on the attached plot?

Happy to add more info if requested, thanks in advance!

Take a look at PCA plot for this dataset.

Usually the above MA plot implies that the majority of the genes are consistent with no difference across the conditions. While the MLEs are scattered around 0, the shrinkage procedure will put a gene with no evidence of a difference at LFC=0.