search

atarashansky / SAMap

SAMap: Mapping single-cell RNA sequencing datasets from evolutionarily distant organisms.
MIT License
64 stars 19 forks source link

Neighbourhood graph or embedding? #106

Closed YY-SONG0718 closed 1 year ago

YY-SONG0718 commented 1 year ago

Dear Alec,

I have a question regarding the outputs of SAMap. Is the final output of SAMap the neighbourhood information stored in sm.samap.adata.uns['neighbors'] and sm.samap.adata.obsp['connectivities'], or the embedding in sm.samap.adata.obsm['wPCA']? Can I use the weighted PCA embedding to calculate neighbourhoods of different sizes?

Thank you so much!

Kind regards, Yuyao

wubaosheng commented 1 year ago

I think the wPCA is the weigh of different species, so it is right to use the wPCA for embeding.

atarashansky commented 1 year ago

Hi @wubaosheng

Please do not use wPCA as an embedding. wPCA is just an intermediate joint embedding used for calculating the stitched neighborhood graph. The correct output is sm.samap.adata.obsp['connectivities'].