summariseScreen: Summarise the drug effect across concentrations

In lujunyan1118/DrugScreenExplorer_dev: A pipeline for processing drug sensitivity screen data

Description

This function summarises the normalized drug effect (viability) across multiple concentrations into a single value. The summarisation is based on the normalized viability values, and therefore plate normalisation needs to be performed first. If the incubation/edge effect corrected viability values are present, the normalisation will be performed on both uncorrected and corrected viability values and a suffix, '.cor' will be added to the columns that contain the summarised values.

Usage

summariseScreen(screenData, method = "average")

Arguments

screenData

the data frame containing screening data generated by readScreen() function

method

a character string or a vector of character strings specifying the methods for summarising the effect. Currently three methods are supported: average, which simply calculates the averaged drug effect across all concentrations; AUC, which calculates the normalized area under dose-response curve using the trapezoidal method; IC50, which will first fits a robust four parameter log-logistic model using dr4pl package and then summarise the drug effect using the four parameters (upper limit, lower limit, slope and IC50/EC50) and area under the sigmoid curve. Multiple methods can be used at the same time by specifying a vector method names.

Value

Depends on the user-specified summarisation method, the following columns will be added to the input data frame:

for method = 'average'

meanViab

the averaged viability across all concentrations for a certain drug

for method = 'AUC'

AUC	the area under the dose-response curve calculated by using the trapezoidal rule.

for method = 'IC50'

UpperLimit	the upper limit of the fitted logistic model
IC50	the half maximal inhibitory concentration (IC50), or the half maximal effective concentration (EC50), depends on the type of input data
Slope	the slope of the fitted logistic model
lowerLimit	the upper limit of the fitted logistic model
AUC	the area under the fitted sigmoid doese-response curve

If the logistic model can not be fitted for a certain drug, the above four values will be set to NA.

If a column that contains the edge/incubation effect corrected viability values is present in the input data frame, the above columns with a suffix .cor, which represent the summarised values based on edge/incubation corrected viabilities, will also be added to the input data frame.

Examples

data(screenData_normalized)
screenData <- summariseScreen(screenData_normalized)
# Please see the vignette for more information.

lujunyan1118/DrugScreenExplorer_dev documentation built on Dec. 21, 2021, 12:42 p.m.