Calibrate model scores with sigmoidal model to approximate interaction probabilities
Usage
X.calibrate.scores(
data = NULL,
standard.set = NULL,
scores.col = NA,
labels.col = NA
)
Arguments
- data
Data frame with columns for protein1, protein2 and a column with scores.
- standard.set
Data frame with column for protein1, protein2 and a column with labels.
- scores.col
Name of the data column with model scores.
- labels.col
Name of the standard.set column with labels with values 0 for non-interacting and 1 for interacting proteins.
Value
A list of three: $data with the ata frame with the original data and an additional column 'probabilities' with the calibrated probabilities,
$plot with the calibration plot and $model with the model fitted into the original data.
Examples
calibrated.model <- X.calibrate.scores(final.model$data, GS)
#> Error in theme_bw(base_size = 12): could not find function "theme_bw"