Scale scores to reference genes

Rescales the scores of a screen so that the median of the non-essential genes is 0 and the median of the essential genes is -1, as done for example by BAGEL2 and Chronos. Scaled scores can be compared between screens with different strength (for example different cell lines or time points): a scaled log2 fold change of -1 means "as depleted as a typical essential gene".

Usage

scale_by_reference(table, essential, nonessential, score = "Log2FC")

Arguments

table

A table with one row for each barcode and the columns Gene and score, for example the output of compute_metrics or compute_barcode_statistics.

essential

The names of the essential genes (as in the Gene column).

nonessential

The names of the non-essential genes.

score

The column with the score (default "Log2FC").

Value

The table with a new column named "Scaled" followed by the name of the score column (for example ScaledLog2FC).

Details

With \(m_N\) and \(m_E\) the medians of the scores of the non-essential and essential genes (gene-level means of the barcode scores), the scaled score of each barcode is \((x - m_N) / (m_N - m_E)\).

See also

compute_screen_quality, hart_reference_genes

Examples

object <- get0("object", envir = asNamespace("ScreenR"))
metrics <- compute_metrics(object,
    control = "TRT", treatment = "Time3", day = "Time3"
)
# Example only: two arbitrary sets of genes used as references
genes <- unique(as.character(metrics$Gene))
scaled <- scale_by_reference(metrics,
    essential = genes[1:20], nonessential = genes[21:60]
)
head(scaled$ScaledLog2FC)
#> [1]   29.78713 -127.87003 -117.81213  -28.35987  299.05660  174.58973