I have an experiment with serial sampling that I want to model (for e.g. presence of disease, disease severity). I want to know what the best way to go about this would be when doing differential expression analyses. My questions are:
- Can I handle repeated measurements using
edgeR? My understanding is no based on previous posts (e.g. Repeated Measures, Within-Subject, Between-Groups RNAseq Analysis ) but I wanted to check this information was up to date and that limma/voom is the best option for this.
- Is there a method that will allow me to use more complex random effect specifications? I understand that limma-voom's
duplicateCorrelationand the dream functions from
variancePartitionwill only allow for random intercepts and not random slopes.