Clustangles: an open library for analysis/clustering of angular data.

Contact

If you have suggestions,
bug reports or
request for improving API/Documentation.

Email: karsar @ ibms.sinica.edu.tw

About

Conformational analysis of macromolecules helps to understand their dynamics and function. Angular data is a valuable source of information for conformational analysis. Clustangles Python package provides methods for analysis and clustering angular data with their application to data obtained from molecular dynamics in mind. Clustangles software is provided under MIT License.

Methods currently included:

Reading data from files:
csv (comma separated values)
output of Amber traj tool for dihedral angles
gromos and charmm outputs for dihedral angles

Tools of Circular Descriptory Statistics: mean, variance, correlation...

Statistical Hypothesis Testing:
Rayleigh tests for uniformity of circular data with specified and non-specified mean direction
tests for a common mean direction
tests for a common median direction
tests for a common distribution

Principal Component Analisys: dPCA, GeoPCA

Clustering:
Hierarchical Clustering (with several types of distances,
corresponding to points in Euclidean space,
on hypersphere and torus)
K-means clustering (modifications as above)
Soft clustering on unit sphere using von Mises-Fisher distributions
Soft clustering on unit sphere with Watson distributions

© 2015 IBMS, Academia Sincia
Template design by Andreas Viklund