Failing when running lamian_test with permutation: Error: $ operator is invalid for atomic vectors

[nickilott]

Hi,

Thanks for developing the tool! I'm just getting started with it but am having a problem running the lamian_test(). I get the following error:

[1] "fitting model: overall: CovariateTest (Model 3 vs.1) or ConstantTest (Model 1) ..."
[1] "The length of fit is ..."
 [1] 1 0 1 0 1 0 1 0 1 0 1
   Mode   FALSE    TRUE 
logical       6       5 
[1] "The length of fit after removing null is ..."
[1] 1 1 1 1 1 1
   Mode   FALSE 
logical       6 
Error: $ operator is invalid for atomic vectors
In addition: Warning message:
In mclapply(seq_len(permuiter + 1), function(i) { :
  scheduled core 1 encountered error in user code, all values of the job will be affected

I am running with the following function call:

tde_test <- lamian_test(
  expr = for_lamian[["expression"]],
  cellanno = for_lamian[["cellanno"]],
  pseudotime = for_lamian[["pseudotime"]],
  design = design,
  test.type = 'time',
  fix.all.zero=FALSE,
  test.method="permutation", 
  ncores=2,
  permuiter = 10,
  verbose.output = TRUE)

where my inputs look like:

for_lamian[["expression"]]

> str(for_lamian[["expression"]])
'data.frame':   7500 obs. of  7047 variables:
 $ EP1_1_AAGGTTCCATGAACCT : num  -0.0119 -0.0494 -0.3792 -0.3093 -0.0119 ...
 $ EP1_1_ACCAGTATCTGTTTGT : num  -0.0119 -0.0494 -0.3792 -0.3093 -0.0119 ...

for_lamian[["cellanno"]]

> str(for_lamian[["cellanno"]])
'data.frame':   7047 obs. of  3 variables:
 $ cell    : chr  "EP1_1_AAGGTTCCATGAACCT" "EP1_1_ACCAGTATCTGTTTGT" "EP1_1_ACCCACTAGCGTCTAT" "EP1_1_ACCCACTCAAGGTGTG" ...
 $ sample  : chr  "GI6642" "GI6642" "GI6642" "GI6642" ...
 $ celltype: Factor w/ 8 levels "Stem/Undifferentiated",..: 4 1 1 1 1 7 1 3 1 4 ...

for_lamian[["pseudotime"]]

>str(for_lamian[["pseudotime"]])
 Named int [1:9811] 1 2 3 4 5 6 7 8 9 10 ...
 - attr(*, "names")= chr [1:9811] "EP1_1_AAGTCTGAGCCCAATT" "EP3_2_CCAGCGAAGTTCCACA" "EP3_1_TTCTACAGTCACACGC" "EP1_2_CTACCCAGTAGAAAGG" ...

design

>str(design)
'data.frame':   8 obs. of  2 variables:
 $ sample : num  1 1 1 1 1 1 1 1
 $ disease: num  2 1 2 1 0 1 0 2

Any help with this issue much appreciated!

Nick

[Winnie09]

Hi! In your data, expression, cellanno, and pseudotime have different numbers of cells (7500, 7047, 9811). This could lead to the expression being empty when running the line expr <- expr[, names(pseudotime), drop = FALSE]. Please try to only use the cells in the intersect of the cells in these three objects. Please let me know if the issue persists. Thanks!

[buriedsand]

I had the same issue. The error happens because for_lamian[["expression"]] is represented as a dataframe instead of a matrix. Convert it to a matrix with as.matrix().