A. Panczakiewicz, V. M. Anisimov, “The Linear Scaling Semiempirical LocalSCF Method and the Variational Finite LMO Approximation”, in Challenges and Advances in Computational Chemistry and Physics, Springer, 2011, Volume 13, Chapter 15, pp. 409-437, Linear-Scaling Techniques in Computational Chemistry and Physics: Methods and Applications, Eds. R. Zalesny, M. G. Papadopoulos, P. G. Mezey, and J. Leszczynski.  author manuscript

---

 

V. M. Anisimov, C. N. Cavasotto, Quantum-Mechanical Molecular Dynamics of Charge Transfer, in Kinetics and Dynamics: From Nano- to Bio-Scale, Eds. P. Paneth and A. Dybala-Defratyka 2010, Vol. 12, pp. 247-266, Challenges and Advances in Computational Chemistry and Physics.  author manuscript

 

20 ps QM MD of ubiquitinn in water dropplet (12199 atoms)

V. M. Anisimov, V. L. Bugaenko, C. N. Cavasotto; ChemPhysChem, 2009, 10 (18) 3194-3196 "Quantum Mechanical Dynamics of Charge Transfer in Ubiquitin in Aqueous Solution" article  supporting materials

---

 

Detailed derivation of the Variational Finite Localized Molecular Orbital approximation is presented. LocalSCF method is explained. 2-layer QM/QM method is developed and applied to protein-ligand docking of p56 LCK SH2 domain (anti-cancer and anti-arthritis target) on a personal desktop computer screening the library of 20,000 drug-like molecules presented by 200,000 bound protein-ligand complexes (with average number of 1700 atoms in each complex).

 

V. M. Anisimov, V. L. Bugaenko; J. Comp. Chem., 2009, 30 (5) 784-798; "QM/QM Docking Method Based on the Variational Finite Localized Molecular Orbital Approximation" article  supporting materials

---

The numerical accuracy of linear scaling semiempirical methods LocalSCF and MOZYME is analyzed in comparison to conventional matrix diagonalization with respect to a variety of molecular properties including conformational energy, dipole moment, atomic charges, and bond orders. Major semiempirical MNDO, AM1, PM3, and PM5 Hamiltonians were considered in the study. As the numerical tests demonstrate, both LocalSCF and MOZYME reasonably reproduce matrix diagonalization results with the deviations being below the accuracy of semiempirical methods. However, the economical LocalSCF memory consumption and faster calculations are more beneficial for the quantum-mechanical modeling of large biological systems. The computational performance of the LocalSCF method is tested on the conformational energy calculation of a series of molecular dynamics snapshots of insulin in a large box of water.

V. M. Anisimov, V. L. Bugaenko, V. V. Bobrikov, "Validation of Linear Scaling LocalSCF Method", J. Chem. Theory Comput., 2006, 2(6) 1685-1692, article  supporting materials