Kernel crushed when running snap.pp.filter_cells

[linlejiang]

Hi! I am processing 18 samples using 50 CPUs and 250G RAM. And one of the samples requires a more stringent tsse filtering criterion. Below is my code, and as I try to applied another tsse filtering criterion on individual sample, the kernel would crush. Any ideas how to fix this?

adatas = snap.pp.import_fragments(
[fl for fl in file_ls],
file=[name for name in file_name_ls],
chrom_sizes=snap.genome.hg38,
min_num_fragments=1000,
sorted_by_barcode=False,
n_jobs=50
)
gff_file='atac/gencode.v41.basic.annotation.gff3.gz'
snap.metrics.tsse(adatas, gene_anno=gff_file)
snap.pp.filter_cells(adatas,
min_counts=2000,
min_tsse=5,
max_counts=100000)

filtered sample 1 with more stringent tsse criterion

snap.pp.filter_cells(adatas[0],
min_tsse=10)