Mouse Whole Cortex and Hippocampus 10x

Technical
, ,
1

It seems that, the provided meta data does not entirely fit the given Count matrix.
First, the meta data contains information of 1169213 cells, while the shape of the matrix (.h5 file) is (1169320 x 31053). So we have a few cells less covered in the meta data.

Consequently the rows cannot be sorted similar to the cells within the count matrix.

I checked it with an older version from the data, which has the correct dimensions for the meta data (1169213 x 31053).
But here I had to sort the meta data accordingly to the cell order in the matrix.

Did I do something wrong using the data?
Or can someone confirm my findings?

If I am correct with the sorting of the meta data, it would be great help to provide the meta data sorted accordingly to the cells in the count matrix :smiley:

Alle the best,
Matthias

2

Hi Matthias,

you are correct that there is a difference in the number of cells included in the matrix and the number of cells in the metadata. Using the sample_name in the metadata you can both subset and order the samples in the expression matrix.

3

Hi Cindy,
Thank you for your reply!
This is how I handled it now.
Are those cells which are not present in the meta data file of low quality or technical artifacts?
Is it worth having a look on them, or can I savely ignore them?

4

The cells that are missing from the metadata are low quality cells and are safe to ignore. Apologies for the confusion that this has caused.

1 Like