Hello,

I am working with data set that has 29 different samples with 9 different conditions (all in Condition column), for a total of 28 contrasts (conditions have 1 to 4 samples).

I am trying to make a PCA plot that shows all samples, but only uses the peaks that are differentially bound in contrast #19 in the PCA calculations. My code is as follows:

dba.plotPCA(proj_contrast_analysis, mask=1:29, label = DBA_CONDITION, labelSize = 0.65, contrast = 19)

Being very new to Diffbind, I would really appreciate if someone could confirm that it is correct. Also, if my code is correct, could someone provide me some advice on how to color each condition separately. Currently the PCA shows all samples, but it uses one color for each of the conditions in contrast 19, and one more color for all other samples (labeled "Other" in the plot).

Thank you very much in advance!

In the call to dba.plotPCA(), set attributes=DBA_CONDITION (instead of label). You can set label to something else if that is helpful.

For example, consider the sample data:

    data(tamoxifen_counts)
    tamoxifen <- dba.contrast(tamoxifen, 
                              group1=3:5, name1="MCF7_Reponsive",
                              group2=8:9, name2="MCF7_Resistant")

    tamoxifen <- dba.analyze(tamoxifen)

    dba.plotPCA(tamoxifen, contrast=1, mask=1:11, 
                attributes=c(DBA_TISSUE), label=DBA_CONDITION)