Mutation effects predicted from sequence co-variation
This package has been superseded by the EVcouplings package, which covers the full computational pipeline including alignment generation, 3D structure prediction, and more. All functionality present in the EVmutation package is also available through EVcouplings; the repository here will only be kept as an archive and not be updated in the future.
Python code to compute mutation effects from a graphical model inferred using plmc. This code requires an up-to-date Python 3 installation, and the following packages (we recommend using the latest Anaconda Python 3 distribution which includes all of these packages by default):
- numpy
- scipy
- pandas
- numba
C code to infer pairwise undirected graphical models for families of biological sequences. For installation instructions, please refer to README.md in the plmc subdirectory. This subfolder contains a copy of plmc which is developed using an independent Github repository.
For a tutorial on how to infer a graphical model from a sequence alignment and how to predict the effects of mutations, please see the included Jupyter notebook EVmutation.ipynb. The repository also includes a static HTML export of the notebook that can be viewed without a Jupyter installation.
If you use this code, please cite the following paper:
Hopf, T. A., Ingraham, J. B., Poelwijk, F.J., Schärfe, C.P.I., Springer, M., Sander, C., & Marks, D. S. (2016). Mutation effects predicted from sequence co-variation. Nature Biotechnology, in press.
The EVmutation module was written by Thomas Hopf in the labs of Debora Marks and Chris Sander at Harvard Medical School. The included plmc code was written by John Ingraham (plmc repository).