README.md

Live cell timelapse apoptosis analysis

The goal of this project is to develop a pipeline to analyze timelapse images of live cells undergoing apoptosis. The pipeline will be able to detect cells, track them over time, and classify them as either apoptotic or non-apoptotic. Each dataset or acquisition condidtions of images each contains timelapse images of HeLa cells in a 96-well plate. These cells are treated with varying concentrations of staurosporine, a drug that induces apoptosis.

Data information

Doses of staurosporine and replicates

Staurosporine concentration (nM)	Number of well replicates
0	3
0.61	3
1.22	3
2.44	3
4.88	3
9.77	3
19.53	3
39.06	3
78.13	3
156.25	3

Image acquisition

For 4 channel data we acquired at the following wavelengths:

Channel	Excitation wavelength (nm)	Emission wavelength (nm)
Hoecsht	405	447/60
ChromaLive 488	488	617/73
ChromaLive 488-2	488	685/40
ChromaLive 561	561	617/73

For 2 channel terminal Annexin V data we acquired at the following wavelengths:

Channel	Excitation wavelength (nm)	Emission wavelength (nm)
Hoecsht	405	447/60
Annexin V	640	685/40

Running the pipeline to extract features from CellProfiler and scDINO

To run the pipeline, first we must install the required packages. We can do this by installing a conda environment. From the base directory of the repository, run the following command:

# change directory to the environment directory
cd environments
# create each of the conda environments needed
conda env create -f scDINO.yaml # for scDINO
conda env create -f timelapse_env.yaml # python environment for the pipeline processing
conda env create -f CellProfiling_env.yaml # for CellProfiler
conda env create -f R_env.yaml # for R and R packages

Do not worry about activating the environments as they will be activated when each shell script is run.

Next, we must run the pipeline. To do this run the main shell script from the base directory of the repository:

source run_full_pipe.sh

The following analysis will be performed: 0. Download the data

1. Preprocess the data
2. Run illumination correction 3a. OPTIONAL: Run CellProfiler optimization of pipeline [Deprecated] 3b. Run SAM2 to track objects over time
3. Run CellProfiler to extract features
4. Process the CellProfiler output
5. Run scDINO extract Deep Learning features
6. Harmonize the features from CellProfiler and scDINO