What should I use for chrom_sizes?

[st4302]

Hello,
I am a beginner with ATAC data and I have a question about the command below:

data = snap.pp.import_data(
    fragment_file,
    chrom_sizes=snap.genome.hg38,
    file="pbmc.h5ad",  # Optional
    sorted_by_barcode=False,
)
data

What should I use for chrom_sizes if I have used the mm10 reference from here (https://support.10xgenomics.com/single-cell-atac/software/release-notes/references#mm10-2020-A-2.0.0) to align my data?

Thank you

[kaizhang]

You can use chrom_sizes=snap.genome.mm10 or arbitrary sizes like chrom_sizes={chr1: 2343, chr2: 2000}. See the documentation for more details: https://kzhang.org/SnapATAC2/api/_autosummary/snapatac2.pp.import_data.html#snapatac2.pp.import_data

[st4302]

Thank you for the help. I also have a question about importing the h5 files.
The cellranger atac output has filtered_peak_bc_matrix.h5 and filtered_tf_bc_matrix.h5. Do I need to import to snapatac2 only the filtered_peak_bc_matrix.h5 matrix?

[kaizhang]

We do not recommend the use of either the filtered_peak_bc_matrix.h5 or filtered_tf_bc_matrix.h5 from the Cell Ranger pipeline. The peaks are called on pseudo-bulk and may miss peaks from rare populations and TF scores generally do not work well on downstream analysis. You should use the fragment file as input to SnapATAC2.

[st4302]

Thank you so much for the fast reply. I will use the fragments.tsv.gz then.