Figure 3

The set of active genes is robust to read depth changes. (a) RNA-seq of CD4+ memory T cells at different read depths. Deeper sequencing increases the fraction of total mass in the low-expression region of the log₂(FPKM) distribution. (b) The Gaussian fit captures the read depth-induced variation in expression distributions. For gene expression distributions at each read depth (x-axis), the Gaussian fit was calculated and the mean plotted (y-axis) vs. read depth. Dashed line, linear regression of log(read depth) vs. mean of Gaussian fit. (c) As read depth increases (x-axis), the cumulative number of detected genes increases (solid line, left axis). However, the fraction of newly detected genes which are expressed in the active region drops (dashed line, right axis). Beyond 24 Mmr, only 58 new active genes are detected.