A unified workflow for data-driven precision cell fate engineering via highly multiplexed gene control
Using pip:
Ensure you have Python and pip installed on your system. Then run the following command:
pip install -r model/requirements.txt
Using conda:
Ensure you have Anaconda or Miniconda installed on your system. Create a new conda environment (optional but recommended):
conda create --name cell_reprogram_env
conda activate cell_reprogram_env
Install the dependencies:
conda install --file model/requirements.txt
Worfklow for processing new dataset and running the model
workflow = precice(adata=adata,
dir='./workflow_dir',
path='./hesc_dataset.h5ad')
## Add cell type transition information
workflow.set_transition(label_map, colname='celltype')
workflow.save_seurat()
## Run differential expression externally in R or internally in Python
workflow.get_DE(DE='../Data/DE/DE_hesc_dataset.csv')
workflow.get_network(cell_type='embryonic stem cell')
## Set up PySCENIC for given dataset
workflow.set_up_pyscenic(species)
## Run PySCENIC (Takes several hours)
workflow.run_pyscenic()
## Post processing of learnt transcriptional network
workflow.learn_weights()
transition = source_name +'_to_' + target_name
workflow.run_precice(species='human',
network_path=self.network_path,
DE_path=workflow.DE_filenames[transition])
## Plot ranked list of perturbations and relative perturbation magnitude
workflow.perturbation_plot(k=15)
## Plot transcriptional network with applied perturbations
workflow.network_plot()