Dear all,
I have realised that when doing exon usage analysis through limma and diffSplice, after filtering exons with more than, let's say, 1 cpm in a number of samples, all the DGEList, or fit objects contain only exons passing that filter. That is OK for the fitting and the testing analysis and results. However, when then used the 'ex' object with the diffSplice results to plot exons within genes with log2FC of Differential usage and pvalues, it plots only exons passing the filter and hence it is lost the information on gene structure to graphically locate exons with significant differential usage.
Could it be possible to use the whole set of exons to plot the results of diffSplice and topSplice that had been done after a cpm filtering?
Thanks
Jose