View source: R/perform.scrublet.R
perform.scrublet | R Documentation |
Removes doublets from dataset.
perform.scrublet(counts = counts, expected_doublet_rate = 0.025)
counts |
Counts matrix |
total_counts |
Total number of cells. NULL = automatically counts. |
sim_doublet_ratio |
Number of doublets to simulate relative to observed |
n_neighbors |
Expected number of neighbours per cell |
expected_doublet_rate |
Expected percentage of doublets to be present in the dataset |
stdev_doublet_rate |
Uncertainty in expected doublet rate |
random_state |
Random state for doublet simulation, approximate nearest neighbour search, nd PCA/Truncated PCA |
synthetic_doublet_umi_subsampling |
Sampling rate for UMIs in a cell when synthesising doublets |
use_approx_neighbors |
Use approximate nearest neighbor method '(annoy)' for the KNN classifier |
distance_metric |
Define distance metric for nearest neighbour calculation: 'angular', 'euclidean', 'manhattan', 'hamming', 'dot'. |
get_doublet_neighbor_parents |
return the transcriptomes of the parent cells for simulated doublets |
min_counts |
Minimum counts per cell |
min_cells |
Minimum number of cells per gene |
min_gene_variability_pctl |
Variability cutoff when deducing highly variable genes prior to PCA reduction |
log_transform |
Log transforms the data |
mean_center |
Should the dataset be centred around the mean |
normalize_variance |
Should the genes have a total variance of 1 |
n_prin_comps |
Number of principal components to retain |
svd_solver |
Character. Which SVD solver to use: 'auto', 'full', 'arpack', 'randomized'. |
print.plot |
Logical. Should doublet plots be printed ? Default = FALSE |
verbose |
Logical. Should function information be printed to hte terminal? Default = FALSE |
seed |
Numerical. What seed should be be set. Default = 1234 |
save.plot |
Boolean. Should the automatically genewrated plot be saved? Default = TRUE |
Doublet-omitted sparse matrix
object <- perform.scrublet(counts = counts)
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.