infercnvpy.pl.chromosome_heatmap()

[yitengfei120011]

Description of feature

How can one draw a chromosome_heatmap that only includes the chromosomes of interest? For instance, if I only draw the graphs of chr1 and chr7, how can I solve this problem?

[grst]

Have you tried subsetting your adata before plotting, something like

adata_sub = adata[:, adata.var["chromosome"].isin(["chr1", "chr7"])]

cnv.pl.chromosome_heatmap(adata_sub, ...)

[yitengfei120011]

First of all, thank you for your reply. I have tested this method before, but I still managed to draw all the chromosomes. My infercnvpy.version is 0.5.0 ,What's the reason for this?

[grst]

Sorry, my bad. infercnvpy() writes the result to .obsm since it does not have exactly the same number of features (since several genes are merged into bins, and some genes are lost at the margins). .obsm is not sliced when subsetting the variable dimension of AnnData.

I'm afraid there's no "nice" way of subsetting the plot then. You could either try to manually subsetting .obsm["X_cnv"] based on the information in .uns["cnv"]["chr_pos"] (and update the .uns["cnv"]["chr_pos"], too, to contain only the regions of interest).

What might be easier, if you can afford the additional runtime, is to subset the AnnData object as discussed above before running tl.infercnv again and redo the plot based on the result.

[yitengfei120011]

Hello, may I ask if it's possible to display the scale of chromosomes on the horizontal axis of the heat map? For example, draw chromosome 1 in the heat map and display the scale of the chromosome on the horizontal axis of the graph.

[grst]

What do you mean by "scale"? length in base pairs?

[yitengfei120011]

May I ask what the horizontal coordinate in the picture represents? Is it the length of chromosome 8? If so, could all of them be drawn?

[grst]

Unfortunately, the x-axis is currently not proportional to the length of the chromosome, but rather the number of measured genes on that chromosome. The unit of the x axis is "bins", where each bin is the rolling average gene expression. For more details, please refer to the description of the infercnv metho.

See also #18.