0
0
Entering edit mode
fdumetz • 0
@32f95c00
Last seen 7 weeks ago
United States

Hello everybody,

Last year I started using SCRAN for scRNA-seq analysis and the person who taught me left. I have a new dataset to analyse. I did the required Rstudio and R updates and then reinstalled SingleCellExperiment and all necessary packages were loaded. please don't shoot me, I have very little knowledge in R but I'm really trying here is my issue

sce <- read10xCounts("/local/..../10X_aggr/outs/count/filtered_feature_bc_matrix")


and I get this error from R

Error: BiocParallel errors
1 remote errors, element index: 1
0 unevaluated and other errors
first remote error:
Error in .check_for_compressed(barcode.loc, compressed): cannot find '/local/.../10X_aggr/outs/count/filtered_feature_bc_matrix/barcodes.tsv' or its gzip-compressed form


The 3 files are here, I tried unzipping and trying again but I always get the error. Can somebody help me please? Thank you in advance

0
Entering edit mode

The file cannot be found.

Either the file really does not exist, or you do not have permissions to read the file.

First, please double check the files really do exist!

Then, please post here what it says if you run from the terminal ls -lh /local/.../10X_aggr/outs/count/filtered_feature_bc_matrix from your terminal - filling in the section you have presumably blocked out

0
Entering edit mode

Are the /..../ in your paths something you put in when you posted, to obscure where you have your data stored? If so, that is not necessary. Knowing where your data are stored on a computer that I have no other knowledge of does me no good if I am in fact a bad man that wants to get your data. And if I did have access, obscuring just part of the path isn't going to effectively deter me from getting your data.

Anyway, I assume that's the case because it's /..../ in your function call and /.../ in the error.

But if not, do note that neither ... nor .... are a thing as far as linux is concerned, and that is likely the error. If it is just to obscure the path, then you need to do

> dir("/local/<with actual path information in this part>/10X_aggr/outs/count/filtered_feature_bc_matrix")


and provide the output. Because R thinks the file isn't there, and if R says it's not there, it's probably not there.

0
Entering edit mode

sorry it is just a stupidly long path (I create too many folders), I just shortened it. I usually copy paste the path after doing pwd so I don't make mistakes.

The 3 files are present and here is what I get when I look at permissions

[fdumetz@crocotta filtered_feature_bc_matrix]\$ ls -lh /local/projects-t3/SerreDLab-3/fdumetz/Leishmania/CD11b/Leish_CD11b_aggr/outs/count/filtered_feature_bc_matrix
total 7.8M
-rw-r--r--. 1 fdumetz igs 318K Jan  2 22:53 barcodes.tsv.gz
-rw-r--r--. 1 fdumetz igs 255K Jan  2 22:53 features.tsv.gz
-rw-r--r--. 1 fdumetz igs 5.3M Jan  2 22:53 matrix.mtx.gz

0
Entering edit mode

But the path you show in your response is different from above? Previously the path included 10X_aggr/outs/count/filtered_feature_bc_matrix, but in your reply it's Leish_CD11b_aggr/outs/count/filtered_feature_bc_matrix.

Also, please either highlight any code and then click on the button above the dialog box that says CODE, or just put a triple backtick (top left key on a QWERTY keyboard) before and after the block of code. When you are writing in the dialog box there is another box just below that shows what your post will look like. If it looks all jumbled up when you are typing, that's how it will look when you post, and converting things like your call to ls to code will make your post readable (I have already fixed your comment so I could read the output).

0
Entering edit mode

The first path was the one of a dataset I already analysed that I tried as a control and I get the same error. The path I checked the permissions from is the right one

0
Entering edit mode

That's weird. All it is doing is

file.exists("/local/projects-t3/SerreDLab-3/fdumetz/Leishmania/CD11b/Leish_CD11b_aggr/outs/count/filtered_feature_bc_matrix/barcodes.tsv.gz")


And it is returning FALSE.

0
Entering edit mode

indeed, same response but definitely the files are there.

0
Entering edit mode

I found the solution, since that folder is on a server I should have put volumes instead of local on the path.