if cost weren't an issue people should probably have as many replicates as they would for any other kind of experiment. we all know that in biology there's usually a lot of variation between samples in whatever we measure.
I should note that in Cuffdiff 2.0 the issues of library size normalization has gone away. It seems then have adopted the same normalization system that DESeq uses. So now if you run samples with significantly different total read counts there's no normalization issue and you're only left with the question as to whether it makes sense to compare samples with very different read depths at all. Right?
I should note that in Cuffdiff 2.0 the issues of library size normalization has gone away. It seems then have adopted the same normalization system that DESeq uses. So now if you run samples with significantly different total read counts there's no normalization issue and you're only left with the question as to whether it makes sense to compare samples with very different read depths at all. Right?
Comment