I'm currently doing an RNAseq analysis of a time course experiment with six timepoints, two genotype conditions and three replicates each. After STAR alignment and transcript quantification I used the DESeq2 package for count normalization and differential expression anlaysis just as decribed in the vignette with the following code:
dds <- DESeqDataSetFromMatrix(countData=count_matrix, colData=coldata, design= ~ genotype + timepoint + genotype:timepoint) dds <- DESeq(dds, test="LRT", reduced = ~ timepoint+genotype) res <- results(dds)
This all works well. However, if I look at the most signficant genes, I get time course plots such as this:
I don't doubt that there is a significant difference between these two conditions but I don't think they are biologically interesting as I seem to picking up noise here. Would you agree with this opinion? Is there an option to use DESeq2 to find genes with a non-parallel course or do I have to use a different package for that?
Thank you very much for your help! If you need any further information, please let me know.