Error in pheatmap 'Error in hclust(d, method = method) : must have n >= 2 objects to cluster' for selected genes

Question

[aackermann287]

Hi all,

I'm driving myself up a wall with this one. I am looking to make a heatmap for a selected group of differential genes. I successfully created a heatmap of top 500 genes using this data with the following code:

topVarGenes <- head(order(rowVars(vst), decreasing = TRUE), 500)
mat  <- vst[ topVarGenes, ]
mat  <- mat - rowMeans(mat)
anno <- as.data.frame(colData(atacDDS)[c("Group")])
pheatmap(mat, annotation_col = anno, show_rownames=FALSE)

When I try to make a heatmap with selected genes using the following code, I get the below error:

selectGene <- c(  "ID_chr2_33134006_33135625" ,"ID_chr9_115087304_115087749", "..")
dataSub <- mat[rownames(mat) %in% selectGene, ]

pheatmap(dataSub, annotation_col = anno, labels_row =c("ID_chr2_33134006_33135625"="LTBP1", "...")

Error in hclust(d, method = method) : must have n >= 2 objects to cluster

If anyone has any suggestions I'd appreciate them!

[Mensur Dlakic]

This is not a direct answer to your question, but it may be helpful if you are truly "driving yourself up a wall." It seems that you are doing something wrong in selecting a subset of genes, but I can't quite see what it is.

Why don't you make a second table manually that contains only genes of interest, and then use the same code that worked above?

[aackermann287]

I appreciate your answer. I am going to go forward with doing that, I originally wanted to avoid it because of my dislike of organizing gene IDs on csv files, but I think it might be the cleanest way to go about it

Answer 1

[yura.grabovska]

The first thing to check is what dataSub looks like - how many rows and columns and if there's any NAs.