Cufflinks FPKM Confidence Intervals--Sources of Variability

[mrbirdy857]

Hello,

Based on the original Cufflinks paper (Nat. Biotech), it would seem that two types of variability are associated with the FPKM measure for a particular transcript (A_t): (i) count variability for the locus (from X_g) and (ii) isoform fraction variability (from gamma_t). Does anybody know if the confidence interval reported by cufflinks incorporates these two types of variability? Or does it only reflect the uncertainty in isoform fraction (gamma_t)?

Also, is this confidence interval based on the normal distribution?

Thanks for any help.