I would like to test the mapping using the
subread::subjunc aligner, but I'm not sure what feature and attributes to use when adding a gtf file. Usually one summarize the data on the gene level using the
exon parameters of the gtf file.
But I would like to work on the transcripts. Would it than make more sense to summarize the reads on the
exon instead in order to better estimate the differences in transcript levels?
Or just running the normal parameters to get the counts on a gene level and later using the
diffSpliceDGE command on the normalized object (like shown in the user's guide).