scrublet
Identify and predict doublets in single-cell RNA sequencing data by analyzing transcriptomes, enabling accurate downstream analysis with configurable parameters for simulation and detection.
Instructions
Predict doublets using Scrublet
Input Schema
| Name | Required | Description | Default |
|---|---|---|---|
| adata_sim | No | Optional path to AnnData object with simulated doublets. | |
| batch_key | No | Key in adata.obs for batch information. | |
| expected_doublet_rate | No | Estimated doublet rate for the experiment. | |
| get_doublet_neighbor_parents | No | Return parent transcriptomes that generated doublet neighbors. | |
| knn_dist_metric | No | Distance metric used when finding nearest neighbors. | euclidean |
| log_transform | No | Whether to log-transform the data prior to PCA. | |
| mean_center | No | Center data such that each gene has mean of 0. | |
| n_neighbors | No | Number of neighbors used to construct KNN graph. | |
| n_prin_comps | No | Number of principal components used for embedding. | |
| normalize_variance | No | Normalize data such that each gene has variance of 1. | |
| sim_doublet_ratio | No | Number of doublets to simulate relative to observed transcriptomes. | |
| stdev_doublet_rate | No | Uncertainty in the expected doublet rate. | |
| synthetic_doublet_umi_subsampling | No | Rate for sampling UMIs when creating synthetic doublets. | |
| threshold | No | Doublet score threshold for calling a transcriptome a doublet. | |
| use_approx_neighbors | No | Use approximate nearest neighbor method (annoy). |