j-andrews7 / ibet Goto Github PK

Genericize to "x-axis" and "y-axis" datasets to minimize width.

Hello @j-andrews7 ,

I followed the README and ran shinyPCAtools, shinyDESeq2, and shinyDECorr. Both shinyPCAtools and shinyDECorr showed amazing interactive views and I will try them on my project datasets.

However, shinyDESeq2 gave me the following error:
Error in DESeq2::lfcShrink(dds, coef = coef, lfcThreshold = lfcThreshold) : coef %in% resultsNamesDDS is not TRUE

I proceeded as follows and got it running:

> resultsNames(dds)
[1] "Intercept"               "cell_N061011_vs_N052611" "cell_N080611_vs_N052611" "cell_N61311_vs_N052611" 
> colData(dds)
> res = results(dds)
> shinyDESeq2(dds, res=res, coef = "dex_trt_vs_untrt", annot.by = c("cell", "dex"))

Even though R documentation for res says "If not provided, it will be generated from the dds object via lfcShrink or results and coef", but shinyDESeq2 did not run without that.

I know its a minor issue, just thought of letting you know in case someone else try it. :-)

Thanks.
Gurpreet

Build in a check,

Initial heatmap annotated properly.

Break out UI and server elements where possible. Move plot functions out of main app code.

Matrix filtering error somewhere.

For easier swapping/comparison between datasets.

This may be why the annotation colors go back to random as well.