I am interested in how exposure to a certain drug affects gene expression levels and splicing in cell lines. We have cell lines derived from over 100 different (human) individuals and have exposed each to control buffer and drug in vitro, so pairing our RNA-seq data when making any differential expression/splicing comparisons is vital. (There is so much variability in expression levels between cell lines that it swamps the drug-induced expression changes, even if every cell line changes in the same direction, if the data are not paired for the comparison.)
I found that outputting variance stabilized data from DESeq and then conducting a simple paired t-test was much more powerful than any other differential gene expression approach I have tried thus far for this dataset (including using cell line identity as a covariate in the generalized linear model.) I would like to try a similar approach using exon-level data to investigate splicing, and was wondering if there is a way to export variance-stabilized exon-level data from DEXSeq, and if so, how? (I am a relative R novice, so example code would be much appreciated.)
If there isn't a way to do this in DEXSeq, and one isn't in the pipeline for being implemented soon, I am wondering if there are any other suggestions for conducting differential splicing/exon analysis without sacrificing the inherent power in our paired dataset?
I found that outputting variance stabilized data from DESeq and then conducting a simple paired t-test was much more powerful than any other differential gene expression approach I have tried thus far for this dataset (including using cell line identity as a covariate in the generalized linear model.) I would like to try a similar approach using exon-level data to investigate splicing, and was wondering if there is a way to export variance-stabilized exon-level data from DEXSeq, and if so, how? (I am a relative R novice, so example code would be much appreciated.)
If there isn't a way to do this in DEXSeq, and one isn't in the pipeline for being implemented soon, I am wondering if there are any other suggestions for conducting differential splicing/exon analysis without sacrificing the inherent power in our paired dataset?