Quality Control

The quality control module enables users to filter samples, cells, and genes, and to detect doublets. A Standard Compute runtime is required.

Workflow	Sample Filtering	Cell Filtering	Gene Filtering	Double Detection	Sample Listing
Microarray	❌	❌	❌	❌	✅
Bulk RNA-seq w/ raw counts	✅	❌	✅	❌	✅
Bulk RNA-seq w/ normalized counts	❌	❌	❌	❌	✅
Single cell RNA-seq w/ raw counts	✅	✅	✅	✅	✅

Sample Filtering

Sample filtering enables users to remove samples based on raw counts or mitochondrial percentage thresholds, so this submodule is not available for Microarray and Bulk RNA-seq with normalized counts datasets.

Sample filtering

Cell Filtering

Remove cells with too few or too many counts, or cells with too high of a mitochondrial gene percentage count.

Gene Filtering

Gene filtering differs between single cell RNA-seq and bulk RNA-seq.

Single cell RNA-seq

For single cell RNA-seq, you need to provide the minimum number of cells that must express the gene (usually 0.1% of cell count).

Bulk RNA-seq

For bulk RNA-seq, you need to provide the minimum number of counts for the genes in a minimum percentage of samples (usually 3 counts in 10% of samples).

Double Detection

Double detection should be used only if a droplet based library preparation kit was used. This command creates a new observation called solo_doublet_prediction with values of singlet or doublet. It is up to the user to eliminate the cell from the Cell Explorer.

Doublet detection is performed using variational autoencoders, namely scVI's SOLO.

Sample List

The sample list submodule offers a quick way for users to see which samples have been added to the analysis and which ones are still in use.

Sample list