Compute per-cell quality control metrics from an ADT count matrix.
Given a feature-by-cell ADT count matrix, this class computes several QC metrics:
- The total sum of counts for each cell, which (in theory) represents the efficiency of library preparation and sequencing. This is less useful as a QC metric for ADT data given that the total count is strongly influenced by the actual abundance of the targeted features, i.e., the presence of a surface protein will typically result in an order-of-magnitude increase to the total count that is independent of the cell's technical quality. Nonetheless, we compute it for diagnostic purposes.
- The number of detected features per cell. Even though ADTs are commonly applied in situations where few features are highly abundant, we still expect detectable coverage of most features due to ambient contamination, non-specific binding or some background expression. The absence of detectable coverage indicates that library preparation or sequencing depth was suboptimal.
- The total sum of counts in pre-defined feature subsets. The exact interpretation depends on the nature of the subset - the most common use case involves isotype control (IgG) features. IgG antibodies should not bind to anything, so high coverage suggests that non-specific binding is a problem, e.g., due to antibody conjugates.
Under the hood, this class is just a pre-configured wrapper around PerCellQcMetrics
.