Sub-Loop Inverse Kinematics Monte Carlo (SLIKMC) is a novel technique for sampling configurations of a kinematic chain according to a specified probability density while accounting for loop closure constraints. This method is an integration of two Markov chain Monte Carlo method, blocked Gibbs sampling and Metropolis-Hastings algorithm and it samples chain configurations in an unbiased manner. SLIKMC is applicable in sampling close-loop/free-end kinematic chains, protein loops in 3D space. The samples generated by SLIKMC are verified to have higher quality than ad-hoc fashion loop construction methods and sampling speed is proven to outrun methods based on discrete search.

SLIKMC is developed by Yajia Zhang and Kris Hauser at Intelligent Motion Lab in Indiana University Bloomington. SLIKMC is implemented using the software package Protein Loop Kinematic Toolkit (LoopTK, https://simtk.org/home/looptk) and released as an extension version of LoopTK.

Download and Installation

Download: package

System requirement: Cygwin/Linux, Cygwin is recommended.

Required software libraries:

1.GNU Scientific Library (GSL) http://www.gnu.org/software/gsl/

2.Mesa 3D Graphics Library (Mesa) version 8.0.2 is recommended. http://mesa3d.sourceforge.net/

3.GLUI User Interface Library http://glui.sourceforge.net/

Installation

1.Install LoopTK package.

For installing LoopTK, please refer to the instruction here: http://ai.stanford.edu/looptk/download.html

2.Install SLIKMC.

Simply go to LoopTK/slikmc and "make". An executable file will be generated with name slikmc.

Contact

For comments or bug issue, please contact the software maintainer Yajia Zhang (zhang.nosp@m.yaj@.nosp@m.india.nosp@m.na.e.nosp@m.du).