pp.import_data()

[pass-c]

Hello, I am newer to analyzing snATACseq data. I am in the initial preprocessing stages and have a very basic question. Some background I am importing fragment files that were generated in from a 10x Multiomic experiment and cellranger-arc.

When I run pp.import_data() the output: AnnData object with n_obs x n_vars = 50540 x 0. In scRNA-seq the n_obs is the nuclei count, does the same go for ATAC? If it is the case and n_obs is cell/nuclei count, could there be an error in the fragment file input (as I know the cell count and the output here does not align with that).

Thank you for your assistance.

[kaizhang]

import_data imports all barcodes/cells. Many of them may not be valid cells. You need to filter them based on number of reads or TSS enrichment.

[pass-c]

Thank you. I have come across your resources for generating snap files. Is this something that needs to be generated from my 10x Genomics cellranger output prior to running snapATAC2 in python? Additionally how do you get the .snap file to a .hdf5 file for downstream analysis? Thank you for your assistance.

[nleroy917]

Can I follow up here. I had some questions too. I'm trying to import some data found on human cell atlas, and eventually I get to a point where it starts throwing some pyo3 exceptions:

pyo3_runtime.PanicException: called `Result::unwrap()` on an `Err` value: Custom { kind: Other, error: "MissingStartPosition: " }

It continues to chew through the fragment files, though. Wondering if it just skips these reads? How concerned should I be about this? It doesn't happen often.

[kaizhang]

@nleroy917 Please check if all files have been downloaded completely using md5sum.

[nleroy917]

Will do ASAP! Thanks I'll let you know how that goes

[nleroy917]

@kaizhang I re-downloaded the data. No checksums I could find, but I dug more deeply, and there were two issues:

1. A few of the fragment files had comments at the top, and
2. Some of them seemed corrupted with random data...

So I will remove those comments and then omit the others for the time being from my analysis. I appreciate all your help and I appreciate the library.