drug_screenings

This module is a companion to the analysis module with functions specifically developped for drug experiments. The common denominator of these functions is that they assume that the metadata contain information about drug type and drug dose and they generally do not consider the possibility of other types of experimenntal variables of interest.

Contents

drug_class

A class designed to handle data from drug screens. The class stores relevant information about the properties of a specific drug and the phenotypic features of the drug replicates at different doses and offers some methods to handle this data for analysis.

Class drugClass - methods available to the user:

• get_mean_doses
• get_dose_dist_from_control
• znormalize
• scale01
• find_const_feat
• remove_constant_feat

filter_compounds

Functions to filter data from a drug screen. Contains a function that removes drug classes with few compounds and functions that detect compounds (drugs) with low effect compared to the controls and filter them out of the dataset.

Functions available to the user:

• remove_MOAs_based_on_drug_count
• compare_drug_to_control_univariate
• compounds_with_low_effect_univariate
• compounds_univariate_effect_sizes
• compounds_with_low_effect_classification

bagging_drug_data

Legacy module that smoothed drug data creating bootstrapped averages. Has been deprecated in favour of the more generally applicable module tierpsytools.analysis.bagging_data.