Typical morphological profiling datasets have millions of cells and hundreds of features per cell. When working with this data, you must
clean the data
normalize the features so that they are comparable across experiments
transform the features so that their distributions are well-behaved ( i.e., bring them in line with assumptions we want to make about their disributions)
select features based on their quality
aggregate the single-cell data, if needed
The cytominer package makes these steps fast and easy.
You can install
cytominer from CRAN:
Or, install the development version from GitHub:
# install.packages("devtools") devtools::install_github("cytomining/cytominer", dependencies = TRUE, build_vignettes = TRUE)
Suggests dependencies may not get installed, depending on your system, so you'd need to install those explicitly.
vignette("cytominer-pipeline") for basic example of using cytominer to analyze a morphological profiling dataset.