Run failed with .csv.gz input file, access denied to output log

How To MapMyCells
,
1

I recently submitted a .csv.gz input file, with cells as rows and genes as columns, to MapMyCells. The run failed (run ID 1771365448267-9bf9ae75-bafb-475c-8529-785359b31b74), and when I tried to download the logs, it opened a new tab with this message (after I disabled my popup blocker):

“This XML file does not appear to have any style information associated with it. The document tree is shown below.

AccessDenied

Access Denied

”

I copied a snippet of the count matrix .csv file below, from before I gzipped it (it is now .csv.gz). I tried re-running it, as this worked for another user, but this also failed and I got the same access denied message on the logs (run ID 1771365448267-9bf9ae75-bafb-475c-8529-785359b31b74 for the second try).

                            MIR1302-2HG FAM138A OR4F5 AL627309.1 AL627309.3

EZ087_seurat_AAACAGCCATAGCGAG.1 0 0 0 0 0
EZ087_seurat_AAACATGCAAGTGTCC.1 0 0 0 0 0
EZ087_seurat_AAACATGCAATAATGG.1 0 0 0 1 0
EZ087_seurat_AAACATGCAATGAATG.1 0 0 0 0 0
EZ087_seurat_AAACATGCACCTGCTC.1 0 0 0 0 0
EZ087_seurat_AAACCAACAAACCTTG.1 0 0 0 0 0
AL627309.4
EZ087_seurat_AAACAGCCATAGCGAG.1 0
EZ087_seurat_AAACATGCAAGTGTCC.1 0
EZ087_seurat_AAACATGCAATAATGG.1 0
EZ087_seurat_AAACATGCAATGAATG.1 0
EZ087_seurat_AAACATGCACCTGCTC.1 0
EZ087_seurat_AAACCAACAAACCTTG.1 0

It’s possible that my .csv.gz file has quotation marks for the column giving the cell names, but a previous post led me to understand that MapMyCells would be configured to allow this.

2

For some reason the first five column names of the .csv file snipped copied in as code, but they should be text.

3

Hi @ddressman91

I acknowledge that the answer I’m about to give you is not very satisfactory.

Failing and then not allowing you to download the logs is generally a sign that the cloud instance running MapMyCells ran out of memory. Crashing because of an “out of memory” error does not give the machine a chance to write any output for download.

Something about your data (~131,000 cells by ~27,000 genes) provoked such a crash. Memory use in MapMyCells is a bit complicated. It is governed by the number of processors assigned to a job, the number of cells each processor is told to map at a time, and the number of marker genes in the taxonomy (this is probably what got you; the Whole Human Brain taxonomy involves more marker genes than Whole Mouse Brain).

Anyway: I downloaded your data, grabbed the first 50,000 cells, wrote them to a CSV file, gzipped it, and was able to successfully map them. I would recommend that you just split up your data into slightly smaller chunks and map each chunk individually (again: 50,000 cells seems fine).

We probably need to do some work on MapMyCells so that it manages memory a little better (dynamically setting the number of cells processed by each processor based on how many genes are in the data). That will take some time. In the meantime: smaller sized inputs are your friend.

I hope that gets you unblocked.

Cheers,

Scott

1 Like
4

That’s fine, thanks for the info. I was going by the Gb limit for input files, but previous to making my full gzipped file, I split the count matrix into 4 parts, so I will probably submit those individually and try again.

5

Hi Scott,

Chunks 1-3 ran fine, but I got an error on chunk 4. I pasted the text of the validation log below if you can’t see it, run ID is 1771449255044-12064d00-2ec9-4c2d-a3f7-0ceb6ad26034. It looks like there’s an NA somewhere (or several of them) that is causing the problem, but I checked the cell names and the count matrix in R and couldn’t find any NA’s. Could NA’s be introduced in the conversion to an .h5ad file or conversion between gene symbols and ENSEMBL IDs?

Log text:

5.10693e-04 seconds == WARNING: Input data is in CSV format; converting to h5ad file at MSA.ctrl.PD.snRNAseq.countmatrix.slice4.csv-2026-02-18-21-19-51.h5ad
7.94722e+01 seconds == an ERROR occurred ====
Traceback (most recent call last):
File cell_type_mapper/validation/csv_utils.py, line 99, in convert_csv_to_h5ad
adata = anndata.io.read_csv(
File anndata/_io/read.py, line 49, in read_csv
return read_text(filename, delimiter, first_column_names, dtype)
File anndata/_io/read.py, line 351, in read_text
return _read_text(f, delimiter, first_column_names, dtype)
File anndata/_io/read.py, line 440, in _read_text
data.append(np.array(line_list[1:], dtype=dtype))
ValueError: could not convert string to float: ‘NA’

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File cell_type_mapper/cli/validate_h5ad.py, line 242, in run
result_path, has_warnings = validate_h5ad(
File cell_type_mapper/validation/validate_h5ad.py, line 97, in validate_h5ad
result = _validate_h5ad(
File cell_type_mapper/validation/validate_h5ad.py, line 136, in _validate_h5ad
write_to_new_path) = convert_csv_to_h5ad(
File cell_type_mapper/validation/csv_utils.py, line 110, in convert_csv_to_h5ad
raise RuntimeError(full_msg)
RuntimeError: =======An error occurred when reading your CSV with anndata:
Traceback (most recent call last):
File cell_type_mapper/validation/csv_utils.py, line 99, in convert_csv_to_h5ad
adata = anndata.io.read_csv(
File anndata/_io/read.py, line 49, in read_csv
return read_text(filename, delimiter, first_column_names, dtype)
File anndata/_io/read.py, line 351, in read_text
return _read_text(f, delimiter, first_column_names, dtype)
File anndata/_io/read.py, line 440, in _read_text
data.append(np.array(line_list[1:], dtype=dtype))
ValueError: could not convert string to float: ‘NA’

Please confirm that your CSV is a table in which each row is a cell and each column is a gene.

7.94722e+01 seconds == CLEANING UP
7.95488e+01 seconds == Mapping algorithm failed because of application errors.
7.95488e+01 seconds == Validation error: e=RuntimeError(‘=======An error occurred when reading your CSV with anndata:\nTraceback (most recent call last):\n File “/usr/local/lib/python3.10/site-packages/cell_type_mapper/validation/csv_utils.py”, line 99, in convert_csv_to_h5ad\n adata = anndata.io.read_csv(\n File “/usr/local/lib/python3.10/site-packages/anndata/_io/read.py”, line 49, in read_csv\n return read_text(filename, delimiter, first_column_names, dtype)\n File “/usr/local/lib/python3.10/site-packages/anndata/_io/read.py”, line 351, in read_text\n return _read_text(f, delimiter, first_column_names, dtype)\n File “/usr/local/lib/python3.10/site-packages/anndata/_io/read.py”, line 440, in _read_text\n data.append(np.array(line_list[1:], dtype=dtype))\nValueError: could not convert string to float: ‘NA’\n\nPlease confirm that your CSV is a table in which each row is a cell and each column is a gene.’), type(e)=<class ‘RuntimeError’>, fname=‘run.py’, lineno=153
Traceback (most recent call last):
File “/usr/local/lib/python3.10/site-packages/cell_type_mapper/validation/csv_utils.py”, line 99, in convert_csv_to_h5ad
adata = anndata.io.read_csv(
File “/usr/local/lib/python3.10/site-packages/anndata/_io/read.py”, line 49, in read_csv
return read_text(filename, delimiter, first_column_names, dtype)
File “/usr/local/lib/python3.10/site-packages/anndata/_io/read.py”, line 351, in read_text
return _read_text(f, delimiter, first_column_names, dtype)
File “/usr/local/lib/python3.10/site-packages/anndata/_io/read.py”, line 440, in _read_text
data.append(np.array(line_list[1:], dtype=dtype))
ValueError: could not convert string to float: ‘NA’

During handling of the above exception, another exception occurred:

Traceback (most recent call last):
File “/apps/run.py”, line 153, in run
runner.run()
File “/usr/local/lib/python3.10/site-packages/cell_type_mapper/cli/validate_h5ad.py”, line 242, in run
result_path, has_warnings = validate_h5ad(
File “/usr/local/lib/python3.10/site-packages/cell_type_mapper/validation/validate_h5ad.py”, line 97, in validate_h5ad
result = _validate_h5ad(
File “/usr/local/lib/python3.10/site-packages/cell_type_mapper/validation/validate_h5ad.py”, line 136, in _validate_h5ad
write_to_new_path) = convert_csv_to_h5ad(
File “/usr/local/lib/python3.10/site-packages/cell_type_mapper/validation/csv_utils.py”, line 110, in convert_csv_to_h5ad
raise RuntimeError(full_msg)
RuntimeError: =======An error occurred when reading your CSV with anndata:
Traceback (most recent call last):
File “/usr/local/lib/python3.10/site-packages/cell_type_mapper/validation/csv_utils.py”, line 99, in convert_csv_to_h5ad
adata = anndata.io.read_csv(
File “/usr/local/lib/python3.10/site-packages/anndata/_io/read.py”, line 49, in read_csv
return read_text(filename, delimiter, first_column_names, dtype)
File “/usr/local/lib/python3.10/site-packages/anndata/_io/read.py”, line 351, in read_text
return _read_text(f, delimiter, first_column_names, dtype)
File “/usr/local/lib/python3.10/site-packages/anndata/_io/read.py”, line 440, in _read_text
data.append(np.array(line_list[1:], dtype=dtype))
ValueError: could not convert string to float: ‘NA’

Please confirm that your CSV is a table in which each row is a cell and each column is a gene.

6

Hi,

I downloaded your slice4.csv.gz file. For some reason, the last line of your CSV is entirely made up of NA (even the cell label is "NA”). I wonder if R “just knows” to skip that line, which is why it didn’t turn up in your debugging efforts. Anyway: clip that line off and you should be fine.

Cheers,

Scott

1 Like