Ready-to-go Jupyter notebook toolbox for plotting AlphaFold-generated MSAs, per-residue pLDDT, and PAE.
- Output files from AlphaFold
- Python with the following packages:
- os
- glob
- pickle
- json
- numpy
- matplotlib.pyplot
- pandas
- Run your AlphaFold (AF) prediction on your preferred system (e.g. on a HPC)
- Download the output folder to your local drive
- Copy the pathname from the downloaded folder and open the correct notebook for your AF version:
| Version 3 (beta) | Version 2.3.2 |
|---|---|
| AF3Plot.ipynb | AFQuickPlot.ipynb (or AFParser.ipynb) |
| Other AlphaFold2 versions may work, but haven't been tested. |
- Run all cells, and paste in the pathname and write in your protein name when prompted.
- (V2.3.2) If the notebook takes too long or the kernal dies, use AFParser.ipynb to create the pLDDT .csv file and go to step 6.
- The plots will be saved inside the output folder as .pdf files, and pLDDT scores as a .csv file
- Use the AFRePlot.ipynb notebook to re-plot pLDDT scores with your choice of nº top-ranked predictions and range of positions shown.
Thank you to Sébastien Lemal PhD for writing the blog post that heavily inspired this Jupyter notebook toolbox: https://blog.biostrand.ai/explained-how-to-plot-the-prediction-quality-metrics-with-alphafold2 A big thank you to the ITS Research Team at QMUL for helping me get my AlphaFold predictions working on our HPC.