theislab/scarches

the difference between scPoli and scANVI

Closed this issue · 1 comments

YawnC commented

Just curious, since they both based on cVAE, with cell-type data as conditioning input, is there any difference between their backbones or algorithms?

moinfar commented

Hi,

Yes.
There are so many differences in the modeling, which results in a significant difference in the model's performance.
I can mention some differences (not all):

  1. scPoli uses prototypes per cell type and employs a metric learning approach to form the latent space.
  2. scPoli maps the covariates in a continuous latent space that reveals sample level variation.
    I suggest reading the paper for more details.