Raw count layer

Batch key registration

HVG selection

Gene filtering

Latent storage key

Normalized layer key

Neighbor graph on latent

library_size parameter

Model saved to disk

uv pip install

K-Dense Web mention

AnnData output saved

Change mode

Delta threshold

FDR filtering

Raw counts input

Batch handling in model

Sort by lfc_mean

Upregulated separation

Downregulated separation

Not Wilcoxon primary

de_summary documents approach

scVI trained first

from_scvi_model init

unlabeled_category set

labels_key in setup

scANVI trained after init

predict() for labels

Predictions stored in obs

Raw counts layer

Batch key registered

TOTALVI model used

protein_expression_obsm_key

Raw counts layer

Batch key registered

Joint latent storage key

Clustering on joint latent

Protein DE requested

Model saved

AnnData saved

summary.md mentions joint modeling

load_query_data used

Reference model loaded

Query model trained

KNN for label transfer

Reference latent as KNN training set

Query latent for prediction

Predictions stored in obs

Query AnnData saved

mapping_log.txt created

PeakVI model used

No layer parameter in setup

Batch key for donors

Latent storage key X_PeakVI

differential_accessibility used

Condition comparison

Reproducibility seed

Model saved

Top peaks exported

Downstream clustering on latent