Defining housekeeping genes suitable for RNA-seq analysis of the human allograft kidney biopsy tissue

BMC Medical Genomics

Table 1 Summary of HKG Datasets Defined in This Study Using 9 Different Normalization Methods

Normalization methods	Expression ratio*							Bias**	Variance**
Normalization methods	0–0.01 (%)	0.01–0.05 (%)	0.05–0.20 (%)	0.20–0.40 (%)	0.40–0.60 (%)	0.60–0.80 (%)	0.80–1.0 (%)	Bias**	Variance**
TC	396 (41.60)	473 (49.68)	78 (8.19)	1 (0.11)	1 (0.11)	0 (0)	3 (0.32)	0.74	0.55
UQ	216 (22.69)	612 (64.29)	115 (12.08)	3 (0.32)	4 (0.42)	1 (0.11)	1 (0.11)	0.45	0.21
Median	157 (16.49)	643 (67.54)	142 (14.92)	7 (0.74)	2 (0.21)	0 (0)	1 (0.11)	0.45	0.22
Quantile	125 (13.13)	655 (68.80)	161 (16.91)	6 (0.63)	3 (0.32)	1 (0.11)	1 (0.11)	0.42	0.18
TMM	236 (24.79)	599 (62.92)	108 (11.34)	4 (0.42)	4 (0.42)	0 (0)	1 (0.11)	0.47	0.23
DESeq	231 (24.26)	610 (64.08)	104 (10.89)	4 (0.42)	2 (0.21)	0 (0)	1 (0.11)	0.43	0.19
TPM	157 (16.49)	643 (67.54)	142 (14.92)	7 (0.74)	2 (0.21)	0 (0)	1 (0.11)	0.45	0.22
RPKM	603 (63.34)	319 (33.51)	26 (2.73)	2 (0.21)	0 (0)	0 (0)	2 (0.21)	1.04	1.03
Lib_size	202 (21.22)	617 (64.81)	123 (12.92)	7 (0.74)	2 (0.21)	0 (0)	1 (0.11)	0.43	0.20

Abbreviations: TC total counts, UQ upper quantile, TMM trimmed mean of M-values, DESeq a differential expression package implemented in R, TPM transcripts per kilobase million, RPKM reads per kilobase per million mapped reads, Lib_size library size
*The expression ratio of each housekeeping gene was calculated by its mean normalized read divided by the maximum reads in its corresponding HKG set
**The bias and variance of each normalization method was calculated by the formulae

ISSN: 1755-8794