I want to use the satuRn package to test for differential transcript usage between two diets. My data is in the sumExp. I set up the contrast matrix L and try to run the satuRn::testDTU function as described in the vignette. I keep on running into the same error and dont unterstand fully where it comes from. Does someone have experience with that can can help?

Thank you so much

1. Count data in SummarizedExperiment 

>sumExp

class: SummarizedExperiment 
dim: 1500 24959 
metadata(1): formula
assays(1): counts
rownames(1500): Gsta3_pericentral Il1r1_pericentral ... Sat1_midzone Tmsb4x_midzone
rowData names(3): isoform_id gene_id fitDTUModels
colnames(24959): AACAGGTTATTGCACC-1 AACCGCCAGACTACTT-1 ... TGTCGTTCATACGATA-18 TGTTCACTGTCTTCCT-18
colData names(6): nCount_RNA nFeature_RNA ... zone sample_name

2.Set up contrast matrix

diet <- as.factor(meta$diet)
design <- model.matrix(~ 0 + diet) # construct design matrix
colnames(design) <- levels(diet)

# Initialize contrast matrix
L <- matrix(0, ncol = 1, nrow = ncol(design)) # One contrast
rownames(L) <- colnames(design)
colnames(L) <- "Chow_vs_HTF"

# Set the contrast for comparing `chow` vs `HTF`
L["HTF", "Chow_vs_HTF"] <- 1
L["chow", "Chow_vs_HTF"] <- -1

> L

     chow_vs_HTF
chow          -1
HTF            1


3. Perform test

sumExp <- satuRn::testDTU(
    object = sumExp,
    contrasts = L,
    diagplot1 = TRUE,
    diagplot2 = TRUE,
    sort = FALSE
)
**
Error in locfdr(zz = zvalues_mid, main = paste0("diagplot 1: ", main)) : 
  ML estimation failed.  Rerun with nulltype=2**