Error sPLS analysis in mixomcis

[Former user]

Hello, I am new in R and trying to use mixomics package for data intergration. However, I have two data matrices (transcriptomics & proteomics datasets) which I want to integrate by implementing sPLS in mixomics including regression mode.

But when i have run the following commands for my datasets:

cratero.pls <- pls(genedataT, proteom, ncomp = 3, mode = "regression") craterospls <- spls(genedataT, proteom, ncomp = 3, keepX = c(100,100,100), keepY= c(100,100,100), mode = "regression")

craterostigma.pls <- perf(cratero.pls, validation = "Mfold", folds = 3, progressBar = FALSE, nrepeat = 10) craterostigma.spls <- perf(cratero.spls, validation = "Mfold", folds = 3, progressBar = FALSE, nrepeat = 10)

It is giving me an error as follows:

Error in solve.default(t(Pmat[, 1:x]) %*% Wmat[, 1:x]) : system is computationally singular: reciprocal condition number = 1.40702e-31 In addition: Warning message: In internal_wrapper.mint(X = X, Y = Y, ncomp = ncomp, scale = scale, : At least one study has less than 5 samples, mean centering might not do as expected

I cannot figure it out. I will be grateful to anyone who could help a beginner like me.

[Al J Abadi]

Hi, sorry we’re getting back to you now as we had not realised we have not updated our issues page on Bioconductor (we’re now on GitHub and discussions are on Discourse).

You probably have highly correlated variables and I can see that you use keepX, keepY arguments so you should use spls function which performs variables selection and can very well deal with correlated variables and not pls.