How to get the actual raw cells behind a metacell?

[chansigit]

Dear authors, thank you for your work. I am wondering if I can get the real cell barcodes behind each metacell so that it would be easier to interprete the metacell's identity. This is important to me because I have already finished cell type assignment on single-cell level and hope to use metacell to enhance my data.

[JulieBaker1]

Dear authors, thank you for your work. I am wondering if I can get the real cell barcodes behind each metacell so that it would be easier to interprete the metacell's identity. This is important to me because I have already finished cell type assignment on single-cell level and hope to use metacell to enhance my data.

Have you solved it? I came across the same problem.

[aviezerl]

Hi @chansigit and @JulieBaker1,

The cell to metacell associations are available at the @mc slot of the tgMCCov object as a named list where names are cell ids (usually barcodes) and values are the metacell associations, e.g.:

mc <- scdb_mc(mc_id)
head(mc@mc) 
#> AAACAGCCAATCCCTT-1 AAACAGCCAATGCGCT-1 AAACAGCCACCAACCG-1 AAACAGCCAGGATAAC-1
#>               44                 22                  7                 24

P.S. We have a new and improved metacell algorithm which is implemented in python at: https://github.com/tanaylab/metacells. We recommend using it instead of the current package.

Aviezer

[UboCA]

Hi @chansigit and @JulieBaker1,

The cell to metacell associations are available at the @mc slot of the tgMCCov object as a named list where names are cell ids (usually barcodes) and values are the metacell associations, e.g.:

mc <- scdb_mc(mc_id)
head(mc@mc) 
#> AAACAGCCAATCCCTT-1 AAACAGCCAATGCGCT-1 AAACAGCCACCAACCG-1 AAACAGCCAGGATAAC-1
#>               44                 22                  7                 24

P.S. We have a new and improved metacell algorithm which is implemented in python at: https://github.com/tanaylab/metacells. We recommend using it instead of the current package.

Aviezer

Hi, I got a lot from metacell this R package. And I'm wondering why you said that python version is more preferential in further time. I know calculating time could be less in python, like scanpy vs Seurat. But how about base algorithm? Is it different between metacell and metacells to some extent?