How to find genes in the same cluster after running "plot_pseudotime_heatmap" command?

[J-sara]

Hello,

I have used "plot_pseudotime_heatmap" to find genes that change as a function of pseudotime. Because I have several genes (around 1000 genes) on the heatmap, the name of the genes that cluster together is not readable. How I can find the name of the genes that cluster together on the heatmap? Using "return_heatmap =TRUE" argument gives an object that includes the gene names on the heatmap but is not still obvious which one are clustered together. I want to perform GO term analysis on the genes that follow a similar pattern during the pseudotime.

Best

[dpcook]

Hey J-sara,

Like you said, return_heatmap creates a pheatmap object, which stores the clustered tree is stored in $tree_row, so you can return the clusters with the cutree function on that: t <- plot_pseudotime_heatmap(x, return_heatmap=T) t <- as.data.frame(cutree(t$tree_row, k=3) colnames(t) <- "Cluster" t$Gene <- rownames(t)

Just be sure to change k=n to match the number of clusters in the heatmap.

Then you can use dplyr::filter on that data frame to get genes associated with each cluster, or you can write it out to a csv and do whatever you want with it

[Xiaojieqiu]

thanks @dpcook !

[honghh2018]

Hi @dpcook The cutree function that set k parameter same as plot_pseudotime_heatmap's num_clusters result showing variable clustering compared with the result of plot_pseudotime_heatmap showing. What happen to this ? and how to fix it ? any advice would be appreciated.