The R package RSEQNORM provides two normalization methods, MIXnorm and SMIXnorm, developed for Formalin-Fixed Paraffin-Embedded (FFPE) tissues RNA-seq data. Note that though designed for FFPE RNA-seq data, MIXnorm and SMIXnorm are applicable to fresh frozen or like sample RNA-seq data. For details, please read the github repository.
install.packages("devtools")
library(devtools)
install_github("S-YIN/RSEQNORM")
library(RSEQNORM)
data(ccRCC)
#Run SMIXnorm and MIXnorm.
#A list is returned.
#$SMIX_normalized_log and $MIX_normalized_log
#are the normalized data in natrual logarithm scale.
#$phi is the proportion of expressed genes.
#$D is the probability of being expressed for each gene.
smix <- SMIXnorm(ccRCC)
mix <- MIXnorm(ccRCC)
normalized.by.smix <- smix$SMIX_normalized_log
normalized.by.mix <- mix$MIX_normalized_log
express.gene.smix <- rownames(ccRCC)[smix$D>0.5]
express.gene.mix <- rownames(ccRCC)[mix$D>0.5]