Some observations and questions ...

[aldojongejan]

Dear Dr. Sauteraud,
I recently started using XCIR (based on your article and the research question of a colleague) and am very grateful for you developing and making the package available!

I have a couple of observations/questions:

• Am I correct that in the code for the 'betaBinomXI' function (in R/betabin.R) the 'p_value' (line 103 & 104) should be changed to 'pvalue' to correspond with line 115 & 116 or vice versa? (Otherwise, I got the error "object pvalue not found")
• In the 'plotQC' function a cross is placed on the position of the 'XIST' gene, but the position is not correct? i.e. as x coordinate Ntrain/2 is used, but shouldn't use 'index' ? (line 778 in R/betabin.R)
• In the 'plotQC' function the 'gene_names' argument/variable is not used and I indeed see in the code it is actually on your 'TODO' list, but you can actually very easily add them using 'ggrepel':

p <- plotQC(bb.cotton[sample == "18-ALD608"], xcig = Cotton_genes_updated)

# if you want to label genes, just add to the vector 'gene_names' (XIST is already marked by a cross, so annotate)
gene_names <- c("XIST")
if (gene_names != ""){
  p <- p + geom_label_repel(data = p$data[GENE %in% gene_names], aes(x = index, y =  fg, label = GENE),
                            label.size    = NA,
                            size = 3,
                            segment.color = 'grey50') 
}
p

This will probably provide you with more control over the labels, then putting it in the function?

with kind regards,
Aldo

PS: I am using the XCIR version 1.7.2