Analyzing miRNA data using MirExpress

[bfantinatti]

Hello guys,
I have to compare two datasets of miRNA sequencing (Illumina data). I installed and ran mirexpress for the two datasets.
Is there any way to compare the results? I intend to observe the miRNAs differentially expressed between these two datasets. Do I have to normalize the data, or the mirexpress output is already normalized?
Tahnk you

[sdriscoll]

Based on the information on their site...I have no idea. One possible clue is whether the "expressions" are integers or if they are floating point values. If they are integer then its likely the are raw, unnormalized counts. If they are decimal values then its possible they have been normalized already. It looks like they deal with raw counts and sums of counts so I'd expect the output to be integer and be unnormalized. One would expect the sum of the counts within a sample to be higher for a sample with more reads than for one with fewer reads. Maybe you can test that theory out to see if the data is unnormalized. One more possible check is to plot the expressions against each other. The points should appear to be mostly symmetrical about a 45 degree line projected from (0,0). If he points are obviously shifted to one side or the other then the data is probably not normalized.

[bfantinatti]

Hello sdriscoll, in fact the values are integers. I tested with other sequence data, and is all the same... Integers.
I saw something about a R package to work with this. I will take a look and post back some informations.
Thank you