中文说明   Download the program   GitLab repo   GitHub mirror   NJU-Git mirror   Documentation

MOKIT offers various utilities and modules to transfer MOs among various quantum chemistry software packages. Besides, the automr program in MOKIT can set up and run common multi-reference calculations in a black-box way.

A list of important utilities along with their functions are shown below

With MOKIT, one can perform multi-reference calculations in a quite simple way, and utilize the best modules of each program. E.g.

UHF(UNO) -> CASSCF -> CASPT2
Gaussian  PySCF  OpenMolcas
or
UHF(UNO) -> GVB -> CASSCF -> NEVPT2
Gaussian  GAMESS  PySCF  PySCF
or
RHF -> GVB -> CASSCF -> ic-MRCISD+Q
Gaussian GAMESS PySCF OpenMolcas

Negligible energy loss(usually<1e-6 a.u., for the same wave function method in two programs) are ensured during transferring MOs, since the basis order of angular momentum up to H(i.e. l=5) are considered.

Note that although MOKIT aims to make the multi-reference calculations black-box, the users are still required to have practical experiences of quantum chemistry computations (e.g. familiar with routine DFT calculations in Gaussian). You are encouraged to learn how to use Gaussian if you are a fresh hand.

Aug 6, 2024

Dependencies on quantum chemistry packages are different for each executable or module. Here the minimum requirements for binary executables automr, frag_guess_wfn and Python modules py2xxx are listed:

1. automr: GAMESS, PySCF
2. frag_guess_wfn: Gaussian
3. Most of the utilities do not depend on quantum chemistry packages except that the modules py2gau, py2orca, py2molpro, etc, work with PySCF installed.

Note that the original GAMESS code can only deal with GVB <=12 pairs. But nowadays we can do hundreds of pairs. To go beyond 12 pairs, please read Section 4.4.10 in manual.

You can choose one of the four options shown below to install MOKIT on Linux or MacOS, and they are for full functionalities. If you only want the utility frag_guess_wfn, see here for an even easier way to install. Pre-built Windows OS executables for a few utilities are provided in Releases. But they are outdated compared with master branch, and there's no way to use full functionality on Windows.

This is the easiest way, but network is required to auto-download the requirements (like Intel MKL). And, creating a new environment before installing is highly recommended, to avoid changing your base environment.

conda create -n mokit-py39 python=3.9 # 3.7~3.11 are available
conda activate mokit-py39
conda install mokit -c mokit

You need to keep mokit-py39 activated when using MOKIT.

If you have no access to network, but still don't want to compile MOKIT manually, you can try option 3.

• Prerequisites:
  • You need to install homebrew on your mac. See here for more tips.
  • You need to install conda via brew and install numpy in base env. via pip, as follows

brew install --cask miniconda
conda init bash # (or zsh) 
conda activate base
pip install numpy

Then brew install ansatzx/homebrew-mokit/mokit

Or brew tap ansatzx/homebrew-mokit and then brew install mokit.

Finally, follow caveats guides, add the following in your zsh(bash/fish etc.) profile.

export MOKIT_ROOT="$(brew --prefix)/Cellar/mokit/master"
export PATH=$MOKIT_ROOT/bin:$PATH
export PYTHONPATH=$MOKIT_ROOT:$PYTHONPATH
export LD_LIBRARY_PATH=$MOKIT_ROOT/mokit/lib:$LD_LIBRARY_PATH

Pre-built Linux executables can be downloaded here.

• Prerequisites: Python3 environment and NumPy.

• A detailed guide for choosing the version of pre-built artifacts and resolving dependencies can be found here

• After downloading the pre-built artifacts, you need to set the following environment variables (assuming MOKIT is put in $HOME/software/mokit) in your ~/.bashrc:

export MOKIT_ROOT=$HOME/software/mokit
export PATH=$MOKIT_ROOT/bin:$PATH
export PYTHONPATH=$MOKIT_ROOT:$PYTHONPATH
export LD_LIBRARY_PATH=$MOKIT_ROOT/mokit/lib:$LD_LIBRARY_PATH
export GMS=$HOME/software/gamess/rungms

The LD_LIBRARY_PATH is needed since the OpenBLAS dynamic library is put there. Remember to modify the GMS path to suit your local environment. Attention: the PYTHONPATH has changed since MOKIT-v1.2.5rc2.

Note that you need to run source ~/.bashrc or exit the terminal as well as re-login, in order to activate newly written environment variables.

The latest version of MOKIT source code can be downloaded here.

• Prerequisites

  • Fortran compiler: ifort(>=2017) or gfortran(>=4.8.5)
  • Intel MKL(recommended) or OpenBLAS
  • f2py (installing Anaconda Python3 recommended)

• Compile all modules

cd src
make all

• After make all, you need to set environment variables MOKIT_ROOT, PATH and PYTHONPATH. E.g. if MOKIT is installed in $HOME/software/mokit, the following should be set in ~/.bashrc:

export MOKIT_ROOT=$HOME/software/mokit
export PATH=$MOKIT_ROOT/bin:$PATH
export PYTHONPATH=$MOKIT_ROOT:$PYTHONPATH
export GMS=$HOME/software/gamess/rungms

Remember to modify the GMS path to suit your local environment. Attention: the PYTHONPATH has changed since MOKIT-v1.2.5rc2.

Note that you need to run source ~/.bashrc or exit the terminal as well as re-login, in order to activate newly written environment variables.

• Each utility is self-explanatory. For example, run fch2inp in Shell, you will find

 ERROR in subroutine fch2inp: wrong command line arguments!  
 Example 1 (R(O)HF, UHF, CAS): fch2inp a.fch  
 Example 2 (GVB)             : fch2inp a.fch -gvb [npair]  
 Example 3 (ROGVB)           : fch2inp a.fch -gvb [npair] -open [nopen]

You can search a utility and read their documentations here.

• For usages of modules in mokit/lib/, see examples/utilities/readme.txt

• The input syntax of the automr program is like Gaussian gjf. For example, the input file 00-h2o_cc-pVDZ_1.5.gjf of the water molecule at d(O-H) = 1.5 A is shown below

%mem=4GB
%nprocshared=4
#p CASSCF/cc-pVDZ

mokit{}

0 1
O      -0.23497692    0.90193619   -0.068688
H       1.26502308    0.90193619   -0.068688
H      -0.73568721    2.31589843   -0.068688

Run

automr 00-h2o_cc-pVDZ_1.5.gjf >& 00-h2o_cc-pVDZ_1.5.out

in Shell. The automr program will successively perform HF, GVB, and CASSCF computations. The active space will be automatically determined as (4,4) during computations. See examples/ for more examples.

• Gaussian
• PySCF
• GAMESS
• OpenMolcas
• Molpro
• ORCA
• BDF
• PSI4
• Dalton
• Q-Chem

• To avoid unnecessary errors, you must specify keywords nosymm int=nobasistransform in Gaussian .gjf file, if you want to provide a .fch(k) file to automr.

• If you find any bug frequently occurs or have any suggestions, you can open an issue on the Issues page.

• You can also contact the developer jxzou via E-mail njumath[at]sina.cn, with related files (gjf, fch, out, etc) attached.

• MOs trasferring among NWCHEM, BAGEL, COLUMBUS, etc.

• Develop/Implement robust black-box strategies of excited state calculations

• Currently we have not published the paper of MOKIT program. If you use (any module or utility of) MOKIT in your work, please cite MOKIT as

  Jingxiang Zou, Molecular Orbital Kit (MOKIT), https://gitlab.com/jxzou/mokit (accessed month day, year)

• If you use MOKIT to peform calculations involving GVB, citing the following two papers would be appreciated

  DOI: 10.1021/acs.jctc.8b00854; DOI: 10.1021/acs.jpca.0c05216.

• If you use MOKIT in your work, please cite MOKIT in the main body of your paper. Citing MOKIT only in Supporting Information of your paper is insufficient. EndNote citation files can be found here. More details and examples of citation can be found in manual. 您的规范引用是对开发者的极大鼓励。您可以使用MOKIT为其他人做计算（包括代算），但务必提醒他/她在发表文章时恰当地引用MOKIT和计算中用到的量子化学软件。

• Click here to see published papers which cited MOKIT.