Xiaojieqiu
commented
6 years ago Hi @sneddonucsf, thanks for pointing out this. actually plot_multiple_branches_pseudotime uses a variance stability transformation (VST) method to transform the data so that is why the scale in plot_multiple_branches_pseudotime is different from plot_genes_branched_pseudotime. We will match them up in our next release to all use log scaling intead of VST.
The warning you are getting is also from the vstExprs function. this warning should not effect the result
Hello!
I've been using the plot_genes_branched_pseudotime and plot_multiple_branches_pseudotime functions to plot the trends of selected genes after a branching point. It appears the scaling for gene expression is different between the two plots. The same gene that ranges from 0-1000 in the plot_genes_branched_pseudotime will be in the range of 0-200 in plot_multiple_branches_pseudotime. Is the data transformed or normalized differently between the two plots? I see there is not a norm_method parameter for the plot_genes_branched_pseudotime.
I'm also getting the following warning message with the plot_multiple_branches_pseudotime that I have not seen with other plotting methods. In sqrt(coefs["asymptDisp"] q (1 + coefs["extraPois"] + ... : NaNs produced
Should I do something to correct this warning or to troubleshoot what the issue may be?
Thank you!