Data and analysis code for Massively parallel protein-protein interaction measurement by sequencing (MP3-seq) enables rapid screening of protein heterodimers
Using Python 3.11.3
- pandas 1.5.3
- numpy 1.24.3
- seaborn 0.12.2
- matplotlib 3.7.1
- scipy 1.10.1
- scikit-learn 1.2.2
- networkx 3.2.1
- graph-tool 2.59 (note that graph-tool has limited OS support, and is only necessary for making the graphs in the DUET figure)
Using R version 4.2.2
- DESeq2 1.38.3
This software has been tested on Windows 11 Home and Ubuntu 18.04.
- benchmarking: Code for comparison of MP3-Seq to benchmark datasets from the paper (Figures 2-3)
- data: Barcode counts for libraries used in the paper (both older and final MP3-Seq method versions)
- design_screening: DUET analysis code for large designed binder screens, and orthogonality gap scripts
- processing_pipeline: MP3-Seq processing pipeline with autoactivator screening, autoactivator correction, psuedoreplication, and R scripts to run DESeq2. The pipeline is shown for the datasets from the paper as an example, along with their expected outputs.
- supplementary_analysis: Misc. scripts to replicate supplementary figures in the paper.
- Alphafold_and_Rosetta_metrics: Scripts to calculate AlphaFold and Rosetta metrics