plot_roc: Plot a ROC curve for a list of protein-protein interactions
In fosterlab/CFTK: The Co-Fractionation Toolkit

Assess the intrinsic quality of a CF-MS dataset by evaluating its ability to recover known protein complexes, using receiver operating characteristic (ROC) analysis. This function visualizes the trade-off between the proportion of true-positive interactions recovered at any given rate of false-positive interactions. In the ideal analysis, this trade-off is skewed towards the top left of the plot area, recovering all true positives without any false positives. The dashed line shows the performance of a theoretical random classifier.

1	plot_roc(pairs)

pairs

a data frame in which each row represents a protein pair, and which contains the following columns:

score: the score assigned to that protein pair, e.g., by a machine-learning classifier, in which higher scores represent a greater probability of a physical interaction
label: a ground-truth annotation of whether that pair of proteins is known to physically interact; one of 1, 0, or NA (not labelled)

a ggplot2 object plotting the ROC curve for the input interactions

fosterlab/CFTK documentation built on Jan. 19, 2021, 10:31 p.m.