Molecular Dynamics simulations of SHAPE experiments with grand-canonical reweighting

Introduction

In this repository you can find code (Jupyter notebooks) and public data associated to this manuscript.

Requirements

The following packages are required to run the notebook:

• barnaba
• mdtraj 1.9.7
• multiprocess
• pandas
• pickle
• python 3.9.12
• scikit-learn
• seaborn

If you manage your dependencies with conda you can use the yml file in the repository.

You can run

conda env create -f environment.yml

to create the conda environment and then

conda activate shapemd

to activate it.

Notebooks

In the notebooks folder you can find Jupyter notebooks that can be run to reproduce most parts of the analyses carried out in the paper. Following is a summary of the notebooks and their usage:

• compute_bindings.ipynb: compute a tensor of binding frequencies from trajectories with different numbers of reagent copies;
• gc_reweight.ipynb: compute the maximum-likelihood partition functions of buffer and binding regions, estimate the chemical potential as a function of the number of particles in the buffer region and compute grand-canonical averages;
• experiments.ipynb: analyse in-house experimental data used in the paper (fit concentration-dependent reactivities and perform statistical tests on GNRA nucleotides.

Example Data

In the data/example_data folder you can find example data. Example data consist of:

• trajectories (.xtc files): a subsampling made with a stride of 250 frames of the original trajectories available for download on Zenodo
• topologies (.gro files): files that can be used as topologies within analyses made with Python package mdtraj
• other: other files containing data extracted from the original trajectories