In the present study, a molecular dynamics study of irinotecan molecule with the atom-centered density matrix propagation scheme was carried out at AM1 semiempirical level of theory, at series of different temperatures, ranging from 5 K to 300 K. Molecular dynamics simulations were performed within the NVE ensemble, initially injecting (and redistributing among the nuclei) various amounts of nuclear kinetic energies to achieve the desired target temperatures. Subsequently to initial equilibration phase of 2 ps, productive simulations were carried out for 8 ps. The accuracy of simulations and the closeness of the generated trajectory to those at the Born-Oppenheimer surface were carefully followed and analyzed. To compute the temperature-dependent rovibrational density of states spectra, the velocity-velocity autocorrelation functions were computed and Fourier-transformed. Fourier-transformed dipole moment autocorrelation functions were, on the other hand, used to calculate the temperature-dependent infrared absorption cross section spectra. The finite-temperature spectra were compared to those computed by a static approach, i.e. by diagonalization of mass-weighted Hessian matrices at the minima located on the potential energy surfaces. Thermally-induced spectral changes were analyzed and discussed. The advantages of finite-temperature statistical physics simulations based on semiempirical Hamiltonian over the static semiempirical ones in the case of complex, physiologically active molecular systems relevant to intermolecular interactions between drugs and drug carriers were pointed out and discussed.
Task-based Parallel Computation of the Density Matrix in Quantum-based Molecular Dynamics using Graph Partitioning
