I am looking for a method or additional software that can make count matrix of genes from human_brain_3k_filtered_feature_bc_matrix.h5. (https://support.10xgenomics.com/single-cell-multiome-atac-gex/datasets/1.0.0/human_brain_3k)
Is there any tool or software? Thanks
•
written
by
yzhao140 •
See Load matrices in Python:
Since these are raw counts you need (probably) to normalize them plus apply quality control etc, so the standard workflows. Check for example the Scanpy package which offers all that.
In case you want to work in R see the Bioconductor Single-Cell Workflow.
Thank you very much! I also have a small question. Is there a way for me to finish the task in laptop via Python. It seems that https://support.10xgenomics.com/single-cell-gene-expression/software/pipelines/latest/advanced/h5_matrices should be done in linux system?
Are you on Windows? I think the entire loading can be done via scanpy, and since it is python it should probably run on Windows, see https://scanpy.readthedocs.io/en/stable/installation.html
But if you search around maybe you find a quality-controlled and normalized count matrix in a suitable format for that dataset, because doing that from scratch without experience takes some time. That probably exists since it is a dataset that 10x distributes.
I am on mac. But https://support.10xgenomics.com/single-cell-gene-expression/software/pipelines/latest/advanced/h5_matrices relies on software "Cell Ranger" which can not be installed in a laptop?
Is this workflow correct? https://cf.10xgenomics.com/supp/cell-exp/notebook_tutorial-3.0.0.html
Log in to answer this question.
What do you mean by "non-programmatic"? What do you want to use it for? A single-cell count matrix is a) large and b) rather meaningless if you cannot "programatically" access it. Excel won't help.
Sorry, there may be a typo here for "non programmatic". I want to get a gene by cell count matrix and use it as input for a model in python.