Hi, I am using the edge package lrt function for a longitudinal microarray study (3 time points: 0,1,3) on 7 individuals (1-7). We are interested in the effect of time on gene expression, using individual time curves, for within-subject comparison. The design is similar to the endotoxin study with 4 subjects, 16 arrays presented in Table 1 of Storey, et al., PNAS 2005.

I built a deSet object with individual:

de_obj <- build_study(data = expr, tme = Interval, ind = Individual, sampling = "timecourse")

And a deSet object without individual:

de_obj <- build_study(data = expr, tme = Interval, sampling = "timecourse")

The difference between the full and the null model is the natural spline fit of the time variable.

full_model <- ~1 + ns(Interval, intercept = FALSE, df = 3)
null_model <- ~1

Using lrt on the deSet ExpressionSet without individual is fine.

de_lrt <- lrt(de_obj)

But an error appears when using lrt on the deSet with individual:

de_lrt <- lrt(de_obj)
    Error in svd(X) : a dimension is zero
    Called from: svd(X)

Since svd(de_obj) runs without a problem, it's unclear how to troubleshoot further.

Two questions: 1) To test the within-subject effect of time on gene expression (similar to the endotoxin experiment with 4 subjects), are these the proper full and null models ? 2) Since svd(de_obj) runs, it's unclear what is causing the svd(X): a dimension is zero error warning. Is there a way to circumvent the error in the code or reformat the input data to run the lrt function with the parameter individual ?

Any input is appreciated. Thank you, Emily

> sessionInfo()
R version 4.1.2 (2021-11-01)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS High Sierra 10.13.3

Matrix products: default
BLAS:   /System/Library/Frameworks/Accelerate.framework/Versions/A/Frameworks/vecLib.framework/Versions/A/libBLAS.dylib
LAPACK: /Library/Frameworks/R.framework/Versions/4.1/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] splines   stats     graphics  grDevices utils     datasets  methods  
[8] base     

other attached packages:
[1] edge_2.26.0         Biobase_2.54.0      BiocGenerics_0.40.0