I selected two GSM samples.

when I use the “champ. QC” function of ChAMP package, a error appears：

> champ.QC(beta = myLoad$beta,pheno=myLoad$pd$type) 
[===========================]
[<<<<< ChAMP.QC START >>>>>>]
-----------------------------
champ.QC Results will be saved in ./CHAMP_QCimages/
[QC plots will be proceed with 411557 probes and 2 samples.]

<< Prepare Data Over. >>
Error in cmdscale(d) : 'k' must be in {1, 2, ..  n - 1}

The complete code is as follows：

pd10 <- data.frame(stringsAsFactors = FALSE,
                  Sample_Name = c("GSM1669564","GSM1669589"),
                  type = c("lung_NSCLC_adenocarcinoma"))
idat.name10 <- list.files("/home/shuangshuang/R/Rstudio/03.MethyICIBERSORT/dataset/LUAD", 
                         pattern = "*.idat") |> substr(1L,30L)
pd10$Sentrix_ID <- substr(idat.name10[seq(1,4,2)],12,21)
pd10$Sentrix_Position <- substr(idat.name10[seq(1,4,2)],23,28) 
pd10$Sample_Type <- ("tumor")
write.csv(pd10,file = "sample_type1.csv",row.names = F,quote = F)
myDir="/home/shuangshuang/R/Rstudio/03.MethyICIBERSORT/dataset/LUAD"
myLoad <- champ.load(myDir, arraytype="450K")
class(myLoad)
#[1] "list"
champ.QC(beta = myLoad$beta,pheno=myLoad$pd$type) 

So how to solve this problem? Thanks!

I think it is because you have only two samples. In the mdsPlot (https://rdrr.io/bioc/minfi/src/R/plot.R) you first calculate the distance between samples, so you obtain a matrix of 1x1 containing the distance between sample1 and sample2. Then, by default mdsPlot calculates cmdscale() but the maximum dimension of the space is by default 2 whereas your distance matrix is of dimension 1 so the error stems from here.

You may create a custom code to make it work for two samples but you may also skip this step with : champ.QC(beta = myLoad$beta,pheno=myLoad$pd$type, mdsPlot=FALSE)