PerturbDecode is a framework developed for the automated analysis of large-scale Perturb-seq screens. The workflow in PerturbDecode consists of:
- Data QC and preprocessing
- Identification of the effects of perturbations on genes with parallel processes
- Learning the regulatory topology of perturbed and impacted genes for single and/or combinatorial perturbations
- Relating the regulatory (genetic) topology to physical interactions
PerturbDecode pipeline consists of a set of Python and R Jupyter notebooks and scripts. Analysis steps which require visual inspection are coded as Jupyter notebooks, while the rest are called from the command line. SRC/Pipeline/parameters.py file contains all the user defined parameters and input/output file names.
Detailed explanations of each step will be included here soon.
SRC/ManuscriptFigures folder contains the analysis scripts developed for generating the figures presented in "Systematically characterizing the roles of E3-ligase family members in inflammatory responses with massively parallel Perturb-seq", Geiger-Schuller, Eraslan et al.