[galaxy-user] How to decide if the deference is significant

Dear All, I am looking for the deferential splicing events between cell types. I have run the Cuffdiff and I am going through the output file "splicing differential expression testing". I have read the documentation and protocols about how Cuffdiff test for differential expression and regulation. However although I know the changes in relative abundance are quantified by the square root of the Jensen-Shannon divergence, I still could not understand the concept of it (unfortunately I am not good at math and statistics). Is there any way to convert "the square root of the Jensen-Shannon divergence" into "fold of diference"? How much of "the square root of the Jensen-Shannon divergence" equals to 2 fold of difference? Thanks. Jianguang