Finding the active genes in deep RNA-seq gene expression studies

BMC Genomics

Table 1 A Gaussian fit describes active genes

Cell line	μ	σ	Threshold	Threshold
Cell line	μ	σ	log₂(FPKM)	zFPKM
GM12878	3.70	1.94	−2.18	−3.03
H1-eSC	3.42	2.18	−1.20*	−2.12*
HMEC	3.77	2.11	−2.37	−2.91
HSMM	3.77	2.05	−2.41	−3.02
HUVEC	3.54	2.27	−1.85	−2.38
HepG2	3.24	2.18	−2.79	−2.77
K562	3.83	1.98	−2.19	−3.04
NHEK	3.45	2.07	−1.96	−2.61
NHLF	3.69	2.07	−2.06	−2.78
Mean +/− SD			−2.11 +/− 0.42	−2.74 +/− 0.30
			−2.23 +/− 0.28*	−2.82 +/− 0.22*

The distribution of log₂ (FPKM) expression for each sample was calculated and the right side of the major peak was fit by a Gaussian distribution with parameters μ and σ. The threshold of active gene expression, defined as the intersection between the linear fit of the active promoter fraction and the repressed promoter fraction, was calculated in log₂(FPKM) and zFPKM. (*) H1 embryonic stem cells were removed as an outlier.

ISSN: 1471-2164