Predicting Cell Health with Morphological Profiles
Gregory Way, Shantanu Singh, Anne Carpenter 2019
The following repository stores a complete analysis pipeline of using Cell Painting data to predict the results of several cell health assays.
Our collaborators previously collected a series of assays measuring different aspects of cell health. The assays include staining with specific dyes to measure specific phenotypes.
We hypothesized that we can use unbiased and high dimensional Cell Painting profiles to predict the readouts of each individual assay.
We collected Cell Painting measurements on a CRISPR experiment. The experiment targeted 59 genes, which included 119 unique guides (~2 per gene), across 3 cell lines. The cell lines included A549, ES2, and HCC44.
About 40% of all CRISPR guides were reproducible. This is ok since we are not actually interested in the CRISPR treatment specifically, but instead, just its corresponding readout in each cell health assay.
We performed the following approach:
- Split data into 85% training and 15% test sets.
- Normalized data by plate (z-score).
- Selected optimal hyperparamters using 5-fold cross-validation
- Trained elastic net regression models to predict each of the 70 cell health assay readouts, independently.
- Trained using shuffled data as well.
- Report performance on training and test sets.
We also trained logistic regression classifiers using the same approach above.
Initial results indicate that many of the cell health phenotypes can be predicted with our approach.
We use conda as a package manager. To install conda see instructions.
To create the computational environment, run the following:
# Make sure the repo is cloned conda env create --force --file environment.yml conda activate cell-health
Each analysis module should be run in order. View each module for specific instructions on how to reproduce results.