I'm in the process of totally re-writing the code, some features on this branch might be temporarily broken

Dependencies for full functionality: NumPy MDTraj Matplotlib SKLearn SciPy MDAnalysis (this will soon be eliminated)

This code contains many functions that may be useful for characterizing intrinsically disordered proteins. Many functions piggyback off of MDTraj. This code has only been tested with Python 2.7.

A major update was just pushed to github and this new version likely contains many bugs. The example at the bottom of "sa.py" should work provided you're running Python 2.7, have correctly formatted input files, and all of the dependencies.

The code can do many things. The central idea behind the code is that IDP's aren't well characterized by rigid geometric descriptors such as secondary structure and tertiary contacts. These two quantities are useful ways to define canonically folded structures, but they don't lend well to structures that have many metastable states and no well-defined folded structure. So, instead, we characterize distributions of states using less-specific criteria such as radius of gyration, asphericity, end-to-end distance, solvent accessible surface area, etc. We can get a lot of leverage out of these quantities if we have enough structures to compute decent statistics.

I also added some membrane analysis functions that may be useful.

So, I mainly use this platform to: 1. Compute Principle Component Analysis to characterize a distribution of structures 2. Compute quantities from polymer theory such as the Flory exponent 3. Analyze the surface chemistry of the protein using the "surface contacts" function 4. Analyze protein-membrane interactions and membrane remodeling

I included several scripts that demonstrate how I use the code in my research (located in the example_wrappers directory). I believe these examples are the easiest way to adapt a project to use this platform. I believe the parallel_rosetta function is the cleanest example.