README.md

In MathurinD/drugResistance: Process growth curves from drug sensitivity experiments

drugResistance

Helper functions to compute drug sensitivity and synergy from incucyte data.

(optional) Load the data using incucyter and check the growth curves.

library(incucyter)
library(drugResistance)
confluency = read_incu(system.file('extdata', 'data.txt', package='drugResistance'), system.file('extdata', 'annotation.csv', package='drugResistance'))
plot_microplate(confluency)

The following columns are needed to compute a growth rate: + Analysis_job: name of the run/experiment + Well: The position of the well as a letter and a number (e.g D5) + Treatment: drug_concentration string separated by a "+" for combinations (eg. AZD6244_10nM+Panobinostat_1nM) + Ref_T: name of treatment used in the control well(s) (drug + conc, e,g DMSO_0.001) + Value: Confluence growth rate + Elapsed: Treatment time (unit is free but should be the same in all the column)

The confluency file from read_incu also has the following columns that are not used in the subsequent functions: - Feature: Type of measurement (e.g. Confluency) - Unit: of the feature (e.g. %) - Metric: Type of mesurement - Reference: position of the control well(s)

Compute growth rates and fit dose-response curves. Growth rates are normalized for each Analysis_job to the average growth of the control wells (specified for each well in column Ref_T)

growth = process_growth_curve(confluency)
plot_growth_rates(growth) # Heatmap of the computed average growth rate in each well
dose_response = fit_drug_sensitivity(growth)

process_growth_curve will add the following columns: - Mean_Well: mean across measurements if the well has several positions (or the only value measured) - Sd_Well: Standard deviation across measurements if the well has several positions (or zero) - Viability: normalised data - Inhibitor: treatement wo/ conc - Concentration: concentration as a string (e.g 20nM) - Concentration_value: concentration in micromolar (e.g. 0.02, you can use the helper function concentration_values to convert from the Concentration column)

Visualise the dose-response curves and export the fitted values.

dose_response$plots
ic50s = data.frame(t(sapply(dose_response, function(xx){c(xx$coef, xx$residual)})))
names(ic50s) = c('slope', 'ic50', 'residual')
write.table(ic50s, system.file('exdata', 'ic50s.csv', package='drugResistance'))

You can also compute and visualise synergy scores (note test data are not synergies, see [Dorel2021][https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1009515] for example data and code)

growth %>% plot_bliss_scores(col_drug="drug1") # Uses the helper function starHeatmap2 which uses ComplexHeatmap to plot growth rate as tile fill color and synergy score as stars

Synergies

If your drug screen has combinations, you can compute synergy:

synergy = get_synergy_table(growth, control='DMSO') # Computes multiple synergy statistics
plot_bliss_scores(growth) # Plots only bliss synergy score
write_csv('synergy.csv', get_synergy_table(growth)) # Can be used as input for https://synergyfinder.fimm.fi/ with minor renaming of the columns

MathurinD/drugResistance documentation built on July 4, 2025, 10:30 p.m.