Unexpected sorting in one column during kmeans clustering

[Mark Manzano]

Hello all,

I was doing k-means clustering on z-scores from gene expression datasets (Job IDs 88665, 88667, 88669, 88670 --- same dataset but with k=3,4,5,6). 

I noticed that the expression values for each cluster are sorted by column 9 for all jobs.

So when I plot the heatmaps, column 9 looks like a smooth gradient while the rest look like a traditional gene expression heatmap (see attached). Col 9 is also a replicate of Cols 7 & 8. 

This seems to hold true when I set my seed differently (I tried seed=1 and seed=12345 for k=6).

Not the end of the world but is there a way to proceed with the clustering but NOT sort by column 9? 

I'd like to get somewhat close to a publication-ready figure and the smooth gradient for 1 replicate is a bit of an eyesore (and I don't seem to see this done in published papers).

Thank you for the help!

Mark

[Barbara Hill]

Hello Mark, 

I'm so sorry we never responded.

Would you still like an answer to this question about sorting in KMeansClustering?

Best

-Barbara