mstool

Multiscale Simulation Tools for Mapping, Backmapping, and Loop Modeling.

Dependencies

python (>=3.7), openmm, sqlite3, cython, pandas, numpy. To install openmm:

conda install -c conda-forge openmm

Installation

Simply run the below for installation.

sh install.sh

The two actions are included in install.sh.

Add PYTHONPATH to the environment (~/.bashrc and ~/.zshrc) and mstool to an environment variable.
```
export PYTHONPATH=$mstoolpath/lib-python:\$PYTHONPATH
export mstool=$mstoolpath/
```

Build a distance matrix, which uses cython.

cd $mstool/lib-python/mstool/lib
python setup.py build_ext --inplace

Tutorial

Please read tutorial.pdf before using mstool.

Mapping scheme

mstool will automatically read the default mapping files: $mstool/mapping/martini.protein.dat and $mstool/mapping/martini.lipid.c36.dat

These are tailored toward martini. Please note that only a few Martini molecules are present in these mapping files at this moment. You should make a new file if you want to add molecules or your coarse-grained structures are not at martini resolution. Under each RESI resname, define a coarse-grained bead in a square bracket ([]) and its corresponding atomistic atoms. If a molecule has any isomers (cis/trans/chiral), define them in a square bracket.

Forcefield

Currently, it only supports charmm36 forcefield. The names of atomistic atoms defined in the mapping schemes should match these defined in the forcefield. All of the standard charmm36 molecules are defined at $mstool/FF/charmm36/charmm36.xml, which will be autmoatically read.

How to use

import mstool
mstool.Ungroup('cg.pdb',  'aa.pdb')
mstool.REM('aa.pdb', 'aa_final.pdb')
mstool.CheckStructure('aa_final.pdb')

cg.pdb is a coarse-grained structure. mstool.Ungroup will randomly place atomistic atoms near their corresponding coarse-grained bead. If you want to add mapping schemes because your molecules are not defined in the default mapping schemes, use mapping_add = ['your_new_mapping1.dat', 'your_new_mapping2.dat']. If you want to overwrite the default mapping schemes, use mapping = ['your_new_mapping1.dat', 'your_new_mapping2.dat'].

REM will energy minimize a system with the modified forcefield. This is an important step as the above step just randomly ungroups atoms. cis/trans/chiral information is read from the default mapping schemes. If you want to add mapping schemes for new molecules, use mapping_add = ['your_new_mapping1.dat', 'your_new_mapping2.dat'].