estimateDispersion: object 'fit' not found

[truittll]

Hello all,

I am trying to run the following lines

cds <- newCellDataSet(as(as.matrix(tpm_mat), "sparseMatrix"),phenoData = pd,featureData =fd, lowerDetectionLimit = 0.5, expressionFamily = negbinomial.size()) cds <- estimateSizeFactors(cds) cds <- estimateDispersions(cds)#having problems

but I keep getting the error message

"Error in glm.fit(x = numeric(0), y = numeric(0), weights = NULL, start = c(1e-06, : object 'fit' not found"

Any ideas what I could be doing wrong? Thank you so much for your help!

[Xiaojieqiu]

I am not aware of this error. you can post a link to your data for me to take a look if you want. thanks

[Kaivalyamolugu]

Hi, I'm facing the same issue. Attaching my data file. Please let me know if you need any other information from me! Thanks! Controldata.txt

[celinesin]

I'm also having the same problem. Just posting here to follow the thread :) Do let me know if I can help in some way!

[Roger-GOAT]

same issue. The cds come from seurat and anndata.

Idents(scRNA) <- 'orig.ident'
subnew <- subset(scRNA,idents=c('day5'))
Idents(subnew) <- 'integrated_snn_res.0.8'
subnew <- subset(subnew,idents=c(3,8,10,12))
sample_ann <- [email protected]  
sample_ann$celltype=subnew$leiden
gene_ann <- data.frame(
  gene_short_name = rownames(subnew@assays$RNA) , 
  row.names =  rownames(subnew@assays$RNA) 
)
pd <- new("AnnotatedDataFrame",
          data=sample_ann)
fd <- new("AnnotatedDataFrame",
          data=gene_ann)
ct=as.data.frame(GetAssayData(subnew,slot = 'counts',assay = 'RNA'))


cds <- newCellDataSet(
  as.matrix(ct), 
  phenoData = pd,
  featureData =fd,
  expressionFamily = negbinomial.size(),
  lowerDetectionLimit=1)

cds <- detectGenes(cds, min_expr = 1) 
cds <- cds[fData(sc_cds)$num_cells_expressed > 10, ]
cds <- estimateSizeFactors(cds)
cds <- estimateDispersions(cds) 

> cds <- estimateSizeFactors(cds)
Warning message:
In if (isSparseMatrix(counts)) { :
  the condition has length > 1 and only the first element will be used
> cds <- estimateDispersions(cds) 
Error in glm.fit(x = numeric(0), y = numeric(0), weights = NULL, start = c(1e-06,  : 
  object 'fit' not found
In addition: Warning messages:
1: In if (isSparseMatrix(exprs(cds))) { :
  the condition has length > 1 and only the first element will be used
2: In glm.fit(x = numeric(0), y = numeric(0), weights = NULL, start = c(1e-06,  :
  no observations informative at iteration 1
3: glm.fit: algorithm did not converge 

[schnappi-wkl]

I'm also having the same problem with truittll

[CJSCI666]

i also having the same problem with truittll

cds <- estimateSizeFactors(cds) Warning message: In if (isSparseMatrix(counts)) { : the condition has length > 1 and only the first element will be used

cds <- estimateDispersions(cds) Removing 69 outliers Warning message: In if (isSparseMatrix(exprs(cds))) { : the condition has length > 1 and only the first element will be used

[unlsysbioywl]

I also have the same problem:

cds <- estimateSizeFactors(cds) Error in if (isSparseMatrix(counts) >= 1) { : the condition has length > 1

Then I did the following steps and finally sovled it.

step1: download the sorce code package from https://www.bioconductor.org/packages/3.15/bioc/src/contrib/Archive/monocle/

step2: decompress the compressed package

step3: type ".libPaths()" in R console to check where the library are

.libPaths() [1] "D:/Program Files/R/R-4.2.2/library"

step4: goto this director and find a directory of the same name "monocle"

step5: replace this directory with the one you just decompressed

step6: reload the sorce code package (maybe you need to install "devtools" first) devtools::load_all("D:/Program Files/R/R-4.2.2/library/monocle")

step7: re-run "estimateSizeFactor(cds)"; it will return the error message and also "Return with Debug" button

step8: click "Return with Debug" button. It will show you at which line in which R script, the error comes from. Or, to be simple, you need to edit utils.R and expr_models.R

step9: edit utils.R at line 35: chnage "isSparseMatrix(cellData)" into "isSparseMatrix(cellData)[1]" at line 192: change "isSparseMatrix(exprs(X))" into "isSparseMatrix(exprs(X))[1]" at line 438: change "isSparseMatrix(counts)" into "isSparseMatrix(counts)[1]" all in all, it is because the return of isSparseMatrix() shoulde return one logical value-TRUE/FALSE but now it return two logical values and both them are FALSE !!!remember to save the changes

step10: edit expr_modles.R at line 454: change "isSparseMatrix(exprs(cds))" into "isSparseMatrix(exprs(cds))[1]" !!!remember to save the change

step11: reload monocle devtools::load_all("D:/Program Files/R/R-4.2.2/library/monocle")

Now, the problem is sovled

cds <- estimateSizeFactors(cds) cds <- estimateDispersions(cds) Removing 159 outliers