Installing ms_deisotope¶
ms_deisotope
was written to be an fusion between a simple peak picking
library, ms_peak_picker and
a simple isotopic pattern generation library brainpy.
It later grew a data access layer written directly on top of pyteomics and lxml.
It has two optional dependencies: Matplotlib for drawing spectra, and psims for writing mzML and updating its compiled controlled vocabullary.
Building From Source¶
ms_deisotope
is written for direct use from Python, but large
chunks of its internals have optional but highly recommended C implementations
that significantly improve the speed of the provided algorithms. To build the
library from source with the C extensions included, you’ll need:
A C compiler matching your Python version
A recent version of Cython >= 0.27.0
NumPy and SciPy
ms_peak_picker built with C extensions
brainpy built with C extensions
Note
ms_deisotope
’s Python implementations and C implementations should
behave in the same way. If you think they are not, please
open a bug report and
let us know!
Once you have all of these dependencies installed, running python setup.py install
should take
care of the rest. If you want to profile the C components, the --include-diagonistics
will enable
profiling hooks.
Installing From PyPI¶
The Python Package Index includes source distributions for ms_deisotope
that work with Python 2.7
and Python 3. Wheels are available for newer version of Python on Windows where C compilers are less common.
To make sure a source distribution build includes C extensions, you must have numpy
, brainpy
(brain-isotopic-distribution
) and ms_peak_picker
compiled and installed first so that their header
files are available.
pip install numpy
pip install -v brain-isotopic-distribution ms_peak_picker
pip install -v ms_deisotope
PEP 517/518 should simplify this process, and this is under active investigation.