Filter a QFeatures assay to samples with complete group annotations

Subsets a QFeatures object to the samples belonging to a specified assay, then removes any samples that have NA in one or more of the given colData columns. Both the assay matrix and colData are filtered consistently.

Usage

filter_complete_groups(obj, i, group_cols)

Arguments

obj

A QFeatures object.

i

Integer or character. Index or name of the assay to filter.

group_cols

Character vector. Column names from colData(obj) to check for NA values. Samples with NA in any of these columns are removed.

Value

A list with the following elements:

mat

Numeric matrix (features x samples) with NA-group samples removed.

cd_assay

Data frame of colData rows corresponding to the retained samples, in the same column order as mat.

removed

Character vector of sample IDs that were removed due to NA in a group column. Zero-length if none were removed.

Details

This is a utility used internally by mnar_score() and mnar_global_score(), but is exported for use in any context where you need a clean sample set with complete group annotations — for example, before building a design matrix or running differential abundance testing.

See also

mnar_score, mnar_global_score

Examples

if (FALSE) { # \dontrun{
# Remove the internal standard sample (NA condition) before modelling
filtered <- filter_complete_groups(obj, "psms_norm", c("Condition", "timepoint"))
mat      <- filtered$mat        # 16 samples, IS removed
cd       <- filtered$cd_assay   # matching colData

# Check which samples were dropped
filtered$removed
} # }