This document discusses statistics and distributions for analyzing RNA sequencing (RNA-Seq) count data. It explains that unlike microarray data, there is no established distribution for RNA-Seq counts based on many replicates. The document explores using the Poisson distribution as a starting point since it models mean and variance equally, but RNA-Seq often shows over-dispersion where variance is greater than the mean. To address over-dispersion, the document considers using the gamma distribution or negative binomial distribution which can combine with the Poisson to account for both technical and biological sources of variation in the counts. It outlines approaches for estimating parameters and correcting for bias to fit an appropriate distribution to the RNA-Seq count data.