Skip to contents

drugsens 0.2.0

New features

  • normalize_to_control(): expresses marker values as a percentage of the control/vehicle group per patient, with multi-marker apoptosis/viability scoring (dead_markers, viable_markers, total_markers, response_metric, consensus_only).
  • plot_dose_response(): plots percent-of-control values against drug concentration using LOESS (method = "loess"), 4-parameter log-logistic Hill curves (method = "4pl" with show_ic50 option, attr(plot, "ic50") parameter table, and multi-marker consensus IC50 attr(plot, "consensus_ic50")), or linear trends (method = "linear"), with optional per-point marker dispersion/CV (consensus_dispersion, show_cv, attr(plot, "consensus_cv")).
  • plot_dose_response_heatmap(): treatment x concentration overview heatmap of percent-of-control with numerical concentration ordering and optional tile value labels (show_values).
  • plot_treatment_summary(): ranked bar chart of mean percent-of-control per treatment with standard error bars and optional statistical comparisons against Control (compare_to_control).
  • plot_patient_heatmap(): patient x treatment overview heatmap of percent-of-control with optional tile value labels (show_values) and statistical comparisons against Control (compare_to_control).
  • plot_normalization_before_after(): side-by-side raw-ratio vs. percent-of-control panels per marker.
  • drugsens_theme() and drugsens_palette: exported publication-ready ggplot2 theme and 8-color CVD-safe categorical palette used across all package figures.
  • Optional statistical comparisons against control (compare_to_control, stat_method, p_adjust_method defaulting to "BH", pvalue_label) in get_QC_plots(), plot_treatment_summary(), and plot_patient_heatmap().
  • Four comprehensive vignettes (drugsens, qupath-and-metadata, quality-control-workflow, cohort-and-dose-response) and a pkgdown documentation website.
  • Expanded testthat unit test suite covering counting, reshaping, control normalization, 4PL dose-response fitting, statistical comparisons, and plotting functions.

Improvements and fixes

  • get_QC_plots() now returns a ggplot object (single patient) or named list of ggplot objects (multiple patients) for interactive and vignette use.
  • Improved label wrapping across all plots so biological acronyms (cCASP3, EpCAM) and concentration units (uM, nM) remain intact.
  • make_run_config() now immediately loads list_of_relabeling after creating config_drugsens.txt on first run.

drugsens 0.1.0

CRAN release: 2025-01-16