These computational methods measure the cell population context in populations of adherent cells from fluorescence images and correct for indirect, population context-determined effects in high-throughput screens as described in Snijder et al., Molecular Systems Biology 2012.

Developed for CellProfiler v1 (in MATLAB). Methods are implemented as CellProfiler modules.

Measure Population Context

% This module calculates the population context parameters (local cell
% density, cell islet edge, cell area and population size), as described by
% Snijder et al., Nature, 2009 and in Snijder et al., Molecular Systems Biology 2012.
% The population context is a set of measurements that describe the spatial
% (and cellular growth) organization of your cells in their images. The
% measurements include: (1) the local cell density; (2) the population
% size; (3) the object (cell or nucleus) area; (4) wether a cell is located at the
% cell islet edge or not; (5) the distance of each cell to the closest edge
% of a cell colony.; and (6) the distance of a cell to its closest
% neighbor.
% If population context is measured per well, the module will also store
% the results of the image filename parsing: (1) Well Row; (2) Well Column;
% (3) Image Site; (4) Microscope type; (5) Timepoint; and (6) the Image
% Snake.

Correct Population Context

% This module calculates the population context-corrected measurement as
% described in Snijder et al., Molecular Systems Biology 2012.
% This module creates a bootstrapped multidimensional bin model of your
% selected measurement against selected population context parameters, and
% corrects your single-cell measurements with the trends described in the
% model. 
% The module will store 3 measurements: (1) "Raw", (2) "Predicted" and (3)
% "Corrected". "Raw" is the raw measurement, "Predicted" the average value
% observed by the bin-model for cells of that population context, and
% "Corrected" the value corrected with the selected method (by default a
% subtraction of raw - predicted).


Publications in which these methods were used:

  1. Mercer, J. et al. RNAi Screening Reveals Proteasome- and Cullin3-Dependent Stages in Vaccinia Virus Infection. Cell Reports 2, 1036–1047 (2012)
  2. Snijder, B. et al. Single-cell analysis of population context advances RNAi screening at multiple levels. Molecular Systems Biology 8, (2012)
  3. Snijder, B. et al. Population context determines cell-to-cell variability in endocytosis and virus infection. Nature 461, 520–523 (2009)