A molecular electronic structure theory program written largely in Python. Jointly developed with the Verstraelen group at Ghent University, Horton is a platform for exploring new ideas, including unconventional electronic structure methods and new conceptual tools for translating computational results into chemical language. Key computational features include geminals-based approaches to strongly-correlated systems, more traditional methods (Hartree-Fock, Kohn-Sham density functional theory, etc.), and the ability to interface to several other quantum chemistry programs. Key interpretative features include density-based conceptual tools and an unusually large variety of population analysis techniques.
A spin- and symmetry-adapted program that optimizes matrix product states using the density matrix renormalization group algorithm. Developed primarily by Sebastian Wouters from Ghent University, CheMPS2 provides an alternative approach to complete- active-space configuration interaction wavefunctions with (CASSCF) and without (CASCI) orbital optimization. By factorizing the CAS wavefunction as a matrix product state, one can roughly double the number of electrons and orbitals that can be treated at the "full CI" level, to about 50 electrons in 50 orbitals. (Parallel versions of CheMPS2 will permit one to treat even larger systems.) This allows one to extend benchmark quantum chemistry calculations to larger systems (more electrons) and higher accuracy (more basis functions). One special feature of CheMPS2 is that it fully exploits the sparse block structure induced by particle number, spin, and abelian point group symmetries.
Exploring Potential Energy Surfaces
The Fast Marching Method (FMM) developed in this group is used to fill up the potential surface by exploring points with the lowest potential value. The algorithm developed expands a wavefront of points to proceed from the reactant to the product. Shepard interpolation, with moving least squares to fit the higher order derivatives of the potential can be used to reduce the number of calculations. The minimum energy path (MEP) can be obtained once the calculation is done. Also we use the quadratic string method (QSM) and the sequential quadratic programming method (SQPM) which start from an initial guess of the MEP and proceed as an optimization algorithm.
A parameterized version of linearized Poisson-Boltzmann solver MEAD (Macroscopic Electrostatics with Atomic Detail). Instructions for compiling the program and running it can be found in the README file in the main directory. Instructions for running para-MEAD can be found in the file examples/README. Many of the PDB files from the paper can be found in this directory.
A program for parameterizing force constants based on the frequencies. This is based on the paper "Automated Parameterization of AMBER force field terms from vibrational analysis with a focus on functionalizing dinuclear zinc(II) scaffolds". The software is available with a tutorial.