How to use the same Kmeans clusters on both original and scaled datasets

[adil-mohamed]

I have used a scaled RNAseq dataset to generate a heatmap and used kmeans clustering to split the rows (genes). Is there a quick and easy way of generating another heatmap using the same kmeans clusters but run on the original non-scaled dataset? ** Thanks for this amazing resource!

[jokergoo]

You can apply k-means before making the heatmap:

km = kmeans(your_matrix, centers = 4)$cluster

Then you can use km to split whatever heatmap:

Heatmap(mat1, row_split = km)
Heatmap(mat2, row_split = km)

[adil-mohamed]

Perfect, thank you!