Multi-sample modality pipeline QC

[TingTingShao]

Dear,

I am following the multi-sample pipeline, after importing the data, I want to see the QC, such as fragment size distribution, tss enrichment score, and num of unique fragments as displayed in the standard pipleine.

%%time
adatas = snap.pp.import_data(
[fl for _, fl in fragments_dict],
file=[name + '.h5ad' for name, _ in fragments_dict],
chrom_sizes=snap.genome.hg38,
# min_num_fragments=1000,
sorted_by_barcode=False,
)

I want to ask if the only way to see this is to display the samples individually.

Thanks
tingting

[kaizhang]

Yes, the built-in plotting functions work on one sample at a time.

[TingTingShao]

Thanks for the quick reply.

So technically, each time when conducting QC filtering, it is done sample by sample?

Could I please ask the reason for this? Why the samples are not combined to generate one plot, the samples are quality controlled based on uniform parameter afterall.

Thanks,
tingting

[kaizhang]

1. Some people may decide to use different cutoffs for different samples.
2. QC is usually done on per-sample basis. Each sample should have its own QC plot.
3. If you want to use a uniform parameter to filter all samples, it can be done easily: just add one line for loop.

[TingTingShao]

I see... Thanks for the explanation.

I saw from the tutorial, one parameter is set for all the samples. Indeed, in real case, it must be more complicated than the example shown in the documentation.

Kind regards,
tingting