We are following the tutorial located here to create a reference and map cell idents to a query. CITE-seq reference mapping with totalVI — scvi-tools
In “Query cell type prediction”, we receive an error that vae_q does not have latent_space_classifier_, when running: predictions = vae_q.latent_space_classifer_.predict(query.obsm["X_totalvi_scarches"]) Should this be vae instead of vae_q? Or do we need to rerun the code to generate the classifier in the “Train a classifier on the latent space” Section? Thank you!