KIF – Key Interactions Finder: A Program to Identify the Key Molecular Interactions that Regulate Protein Conformational Changes

Starting structures, input files, analysis scripts and datasets used corresponding to the article:

KIF – Key Interactions Finder: A Program to Identify the Key Molecular Interactions that Regulate Protein Conformational Changes.

Further details (Same as in the readme file) are provided below:

Note 1. the formatting of this document may look odd if you do not have "word wrap" switched on in your text editor. 

Note 2. Both the MD simulations and PyContact calculations were run on computer clusters running Slurm Workload Manager.
(So if you see any .slm files or commands that start with "sbatch", those relate to that.)

This repo is split into 4 sections: 

######### 1_MD_Simulation_Inputs: #########
The starting structures used to perform all MD simulations in our paper.

Alongside the input files used to run the MD simulations and HREX-MD simulations.

The simulations of PTP1B and the PDZ3 domain were performed with AMBER.

The simulations of the three KE07s were performed with GROMACS.

See the paper for details on how each structure was prepared. 

#########2_MD_Simulation_Analysis: ######### 
Analysis scripts used to calculate various properties with CPPTRAJ.

Also included are the shell scripts used to classify each frame for simulations of PTP1B and each KE07. 

######### 3_Run_PyContact: #########
The shell scripts used to setup, run and collate the PyContact calculations run for each system are provided here. 

Included are the template python scripts (for PyContact) and slurm scripts (used to submit jobs to the computer cluster) used.

The outputs from these calculations (many .csv files) were then processed directly with KIF (our program described in the paper). These outputs are provided in this repository:  "4_Feature_Target_Datasets"

Average run time for an individual run was between 1-2 hours on 16 core CPU. Of course this would depend heavily on the system size and number of frames.   

######### 4_Feature_Target_Datasets: #########
Feature and target datasets used by KIF to generate the data for this manuscript. 

Some of these data sets are also used in the tutorials which are hosted on the GitHub repository for the package. See: https://github.com/kamerlinlab/KIF 

######### 5_Bootstrapping: #########
Files used to bootstrap the datasets (provided in 4_Feature_Target_Datasets) for PTP1B and the PDZ3 domain. 

We have a tutorial on our GitHub https://github.com/kamerlinlab/KIF demonstrating how one can work up these results.