pheatmap input file

Question

[Bioinfonext]

Hello,

Can we use DESeq2 normalise count as input for pheatmap?

x<-counts(dds6Genus,normalized=TRUE)    
heatmap(x,scale="row",color =colorRampPalette(c('blue','yellow'))(12),cluster_cols =F,show_colnames = T,cluster_row=T,border_color="NA", fontsize_row=10,fontsize =10)

Many thanks

Answer 1

[ATpoint]

One usually uses the normalized counts on the log scale, after Z transformation, see:

How to work out Z score for heatmaps for RNA seq dataset

[Bioinfonext]

Is anything wrong in above command as we are taking normalize count and then in pheatmap we are doing scaling using "scale=row"?

[ATpoint]

That is probably the same in terms of Z score. Be sure to input counts on the log scale.

[Bioinfonext]

Thank you so much, so should I used below set of commands?

x<-counts(dds6Genus,normalized=TRUE)  
#For log transformation
x2 <- log2(x)

heatmap(x2,scale="row",color =colorRampPalette(c('blue','yellow'))(12),cluster_cols =F,show_colnames = T,cluster_row=T,border_color="NA", fontsize_row=10,fontsize =10)

Many thanks,

[ATpoint]

No, vst runs on the dds object, not the counts, please see ?vst.

[Bioinfonext]

Sorry,Can we just do log transformation as updated in the code. Many thanks for your quick help.