gravatar for pennakiza

2 hours ago by

Hello everyone,

I have a question re filtering for low variance prior to WGCNA.
I have got RNASeq data, pre-filtered for low counts and transformed with DESeq2 vst. I was wondering if you could help me select from the two methods below the one that is more correct for my data.

filter <- function(x)(IQR(x, na.rm=T)>0.25)
filtered_genes <- genefilter(df,filter)
df_filtered<-df[filtered_genes,]

or

data$variance = apply(data, 1, var)
data = data[data$variance >= quantile(data$variance, c(0.25)), ]
data$variance <- NULL

Thank you very much for your help!

Penny



Source link