In this study, we present some modifications in the semiempirical quantum chemistry MOPAC2009 code that accelerate single-point energy calculations (1SCF) of medium-size (up to 2500 atoms) molecular systems using GPU coprocessors and multithreaded shared-memory CPUs. Our modifications consisted of using a combination of highly optimized linear algebra libraries for both CPU (LAPACK and BLAS from Intel MKL) and GPU (MAGMA and CUBLAS) to hasten time-consuming parts of MOPAC such as the pseudodiagonalization, full diagonalization, and density matrix assembling. We have shown that it is possible to obtain large speedups just by using CPU serial linear algebra libraries in the MOPAC code. As a special case, we show a speedup of up to 14 times for a methanol simulation box containing 2400 atoms and 4800 basis functions, with even greater gains in performance when using multithreaded CPUs (2.1 times in relation to the single-threaded CPU code using linear algebra libraries) and GPUs (3.8 times). This degree of acceleration opens new perspectives for modeling larger structures which appear in inorganic chemistry (such as zeolites and MOFs), biochemistry (such as polysaccharides, small proteins, and DNA fragments), and materials science (such as nanotubes and fullerenes). In addition, we believe that this parallel (GPU-GPU) MOPAC code will make it feasible to use semiempirical methods in lengthy molecular simulations using both hybrid QM/MM and QM/QM potentials.

In general, computational simulations of enzymatic catalysis processes are thermodynamic and structural surveys to complement experimental studies, requiring high level computational methods to match accurate energy values. In the present work, we propose the usage of reactivity descriptors, theoretical quantities calculated from the electronic structure, to characterize enzymatic catalysis outlining its reaction profile using low-level computational methods, such as semiempirical Hamiltonians. We simulate three enzymatic reactions paths, one containing two reaction coordinates and without prior computational study performed, and calculate the reactivity descriptors for all obtained structures. We observed that the active site local hardness does not change substantially, even more so for the amino-acid residues that are said to stabilize the reaction structures. This corroborates with the theory that activation energy lowering is caused by the electrostatic environment of the active sites. Also, for the quantities describing the atom electrophilicity and nucleophilicity, we observed abrupt changes along the reaction coordinates, which also shows the enzyme participation as a reactant in the catalyzed reaction. We expect that such electronic structure analysis allows the expedient proposition and/or prediction of new mechanisms, providing chemical characterization of the enzyme active sites, thus hastening the process of transforming the resolved protein three-dimensional structures in catalytic information.

No abstract

No abstract

Obtaining reactivity information from the molecular electronic structure of a chemical system is a computationally intensive process. As a way of probing reactivity information around that, there exist electron density response variables, such as the Fukui functions (FFs), which are well‐established descriptors that summarize the local susceptibility to react. These properties only require few single‐point quantum chemical calculations, but even then, the intrinsic high cost and unfavorable computational complexity with respect to the number of atoms in the system makes this approach available only to small fragments and systems. In this study, we explore the computation of FFs, showing that semiempirical quantum chemical methods can be used to obtain the reactivity information equivalent to that of a Density Functional Theory (DFT) functional, for the eight entire polypeptide chains. The combination of semiempirical methods with the frozen orbital approximation allows for the obtention of these reactivity descriptors for biological systems with reasonable accuracy and speed, unlocking the utilization of these methods for such systems. These results for the frozen orbital approximation can be additionally improved when other molecular orbitals from the frontier band are employed in the computation. We also show the potential of this computational protocol in the ligand–protein complexes of HIV‐1 protease, predicting which of those ligands are active inhibitors.