Systematic genetic characterization of the human PKR kinase domain highlights its functional malleability to escape a viral substrate mimic
A deep mutational scan of human PKR against the poxvirus pseudosubstrate antagonist vaccinia K3, by Michael Chambers, Sophia Scobell, and Meru Sadhu.
This repo provides supporting code and results this project, a preprint of the results is available here.
We generated a library of single-residue variants in the human protein PKR, then screened this library against a pseudosubstrate antagonist, vaccinia K3. We identified genetic variants that render PKR nonfunctional in the absence of K3, as well as variants that are resistant or susceptible to K3. This repo provides the code and results used for the analysis and figures included in the manuscript. Briefly, our analysis can be divided into three portions: (1) design of primers to generate PKR variants, (2) linking PKR variants to genetic barcodes, and (3) analysis of the PKR library screen against vaccinia K3. Most all figures were generated from the file results/barseq/pkr-variant-reads_240228.csv and Jupyter notebooks for analysis are in the workflow directory.
dms_human_pkr
├── data
│ ├── dms_primers - supplemental files for the design of pkr
│ └── pacbio - supplemental files for PacBio sequencing run
├── dms_primer_design - code to design primers to make PKR variant library
├── results
│ ├── alignparse - output contains codon table linking barcodes to pkr variants
│ ├── barseq - output files of barseq experiment for analysis and figures
│ └── figures - images used for manuscript
├── structure_predictions
│ └── AlphaFold2 - structure predictions used for analysis and images in manuscript
├── supplemental_data - supplemental files for manuscript
│ ├── 1_Oligo-Table.xlsx
│ ├── 2_Variant-Primers.xlsx
│ ├── 3_Plasmid-Table.xlsx
│ └── 4_PKR-Functional-Score-Table.csv
└── workflow - code and environment files used for data analysis and figures
└── notebooks - Jupyter notebooks used for data analysis and figures