| Author: | Gurpreet Singh and Alex Chen |
|---|---|
| Contact: | togurpreet@gmail.com |
| License: | Read LICENSE.txt |
Parallelclusterer is a python module that can be used for clustering 3D objects (proteins) in parallel. Currently only Daura clustering algorithm is implemented.
- Numpy (developed using version 1.8.0)
- PyYaml
- Gromacs (version 4.5.*) developed using 4.5.5
- gp_grompy (uses Gromacs-4.5.5)
- gcc compiler
- gcc-4.7.3 and above. NOTE: The gcc-4.4.* and gcc-4.1.* give troubles
- tables (optional)
- some utilities will not work if is absent
- mpi4py (developed using version 1.3)
- Msmbuilder (optional) version 2.7
Gromacs include and lib directories are needed for compilation of C code.:
python setup.py build_ext --include-dirs < gromacs include directory > --library-dirs < gromacs library directory > python setup.py install
The installation script inst.sh is provided. Please modify it according to your environment.
The script will not work as is.
!DANGER!
The script naively tries to remove the previous installation using rm command, DO NOT USE IT WITHOUT MODIFICATIONS.
In some cases -fexcess-precsion=fast is not supported. Deleting the flag from setup.py should do the trick.
The directory test_dataset0 contains test.sh. The script shows the basic usage. Use cluster.py -h for further help.