Interpreting Overlap QC metrics

[Dongeun Heo]

Hello -

I wanted to ask how to interpret the Overlap QC metrics when running make_cicero_cds.

> cicero_cds <- make_cicero_cds(input_cds, reduced_coordinates = umap_coords)

Overlap QC metrics:
Cells per bin: 50
Maximum shared cells bin-bin: 44
Mean shared cells bin-bin: 0.078793033925032
Median shared cells bin-bin: 0

I saw on the Bioconductor page (https://www.bioconductor.org/packages/devel/bioc/vignettes/cicero/inst/doc/website.html) that On average, more than 10% of cells are shared between paired bins. If I see a much lower value (0.07879), what does that mean? If this means something is wrong, how would I go about fixing/filtering the dataset?

Thank you!

Dongeun

[hpl...@gmail.com]

Hi Dongeun,

The warning that prints on the vignette is because for the vignette the data size is very small (so it runs fast). It's better not to have the warning. Yours looks good to go!

Best,

Hannah