Numerous methods are used in the active research area of simulating quantum many-body systems on diverse computing platforms. Commonly used methods, especially coupled cluster techniques, have been adapted to utilize the most recent breakthroughs in high-performance computing. Configuration interaction (CI) methods, specifically those selected (sCI), have been widely employed and refined in recent years. Still, the refinement of sCI methods designed for massively parallel architectures has been examined in only a small number of scholarly studies. We introduce a parallel, distributed memory implementation of the adaptive sampling configuration interaction approach (ASCI) for sCI. This work will specifically examine the key challenges of parallelizing determinant search and selection, Hamiltonian development, and variational eigenvalue computations within the framework of the ASCI method. Through the application of memory-efficient determinant constraints, originally developed for the ASCI-PT2 method, load balancing in the search step is accomplished. Benchmarks on ASCI calculations for Cr2 (24e, 30o) with 106, 107, and 3 x 108 variational determinants reveal near-optimal speedup across up to 16384 CPUs. This variational ASCI calculation, to the best of the authors' knowledge, is the largest undertaken to date.An exploration of the error sources within the Thomas-Fermi-von Weizsäcker (TFW) density functional, contrasted with the Kohn-Sham density functional theory (DFT), is undertaken. Detailed numerical studies encompassing a variety of strained materials and crystal structures with atomic displacements indicate that, although the ground state electron density in the TFW orbital-free DFT approach closely mirrors the Kohn-Sham density, the corresponding energy calculation significantly diverges from the Kohn-Sham result. The TFW approximation's inadequate representation of the linear response in the electronic kinetic energy, we demonstrate, is the cause of these observed disparities, thereby corroborating existing theoretical predictions. Employing the TFW electronic ground state density within a non-self-consistent Kohn-Sham calculation yields an energy value that closely mirrors the energy outcome of a fully self-consistent Kohn-Sham calculation.The precision of ground-state energies obtained using the transcorrelated Hamiltonian is shown, with the aid of sophisticated Jastrow factors from variational Monte Carlo and further complemented by coupled cluster and distinguishable cluster methods at the single and double excitation levels. Utilizing the cc-pVTZ basis, the transcorrelated distinguishable cluster approach demonstrates near-complete basis set and full configuration interaction quality in predicting relative energies for systems encompassing over thirty atoms and molecules. To evaluate performance under different correlation conditions, we additionally scrutinize the fragmentation of the nitrogen molecule via transcorrelated coupled cluster methodologies. The effectiveness of a streamlined approach for integrating the major effects arising from three-body integrals into amplitude equations is supported by numerical data, obviating the need for their direct introduction.Molecular dynamics often employs sampling strategies, including umbrella sampling and alchemical free energy methods, which necessitate sampling across multiple states. The formalism of the Multistate Bennett Acceptance Ratio (MBAR) is a prevalent approach for the recombination of the resulting data. Previous error analyses of the MBAR estimator have, however, mistakenly assumed independent samples, obscuring a deeper understanding of the error. This paper presents a central limit theorem for MBAR estimates that addresses the complication of correlated data, thereby further supporting the use of MBAR in practical scenarios. The central limit theorem, in addition, yields an estimation of the error which can be segregated into the parts arising from individual Markov chains used for the sampling of states. The impact of state-specific sampling methods on the total error is further elucidated. Through an examination of umbrella sampling calculations for the free energy of alanine dipeptide isomerization and alchemical calculations for the hydration free energy of methane, our error estimator is showcased. The time required for the Markov chain to decorrelate in individual states, as evidenced by our numerical results, plays a considerable role in the overall MBAR error, thereby highlighting the need for accurate consideration of sample correlation.We adapt a density matrix-based optimization technique to design custom basis functions capable of representing chains of rotating water molecules under various intermolecular distance-dependent interaction scenarios. The procedure yields a very condensed basis, whose truncation point is explicitly tied to the population counts of its constituent single-particle basis functions. For the water trimer, we investigate the convergence trends of its diverse properties, showcasing a demonstrably superior performance in comparison with an energy-based truncated basis. Demonstrating the efficiency of the optimized basis, the number of required basis functions is decreased by at least a factor of ten. Ultimately, the optimization process is used to examine longer chains of up to ten water molecules. Hydrogen bond formation and its impact on the net polarization of the chain are considered.A recent study in the Journal of Chemical Physics, authored by J. A. Tan and K. U. Lao, presents the Grassmann interpolation (G-Int) method. Exploring the concepts of physics. Extension of 158, 051101 (2023) to spin-unrestricted open-shell systems has been achieved. Closed-shell systems benefit from a single G-Int application, since their alpha and beta density matrices are identical; conversely, spin-unrestricted open-shell systems necessitate performing G-Int twice, once for each spin density matrix (alpha and beta). In this study, we evaluated the performance of G-Int on the carbon monoxide radical cation (CO+) and the nickelocene complex, possessing doublet and triple ground states, respectively. The Frobenius norm errors for the interpolated spin and density matrices displayed similar magnitudes for the same molecular configuration. In self-consistent field (SCF) calculations, the use of G-Int density matrices as initial guesses yields superior results compared to conventional SCF guess methods, specifically superposition of atomic densities, purified superposition of atomic densities, the core Hamiltonian, and the generalized Wolfsberg-Helmholtz approximation. The desired accuracy of the calculation dictates the use of G-Int density matrices to directly evaluate the self-consistent field (SCF) energy without the computational overhead of iterative SCF procedures. Furthermore, the spin-unrestricted G-Int density matrices were utilized for the initial calculation of atomic charges, employing the Mulliken and ChElPG population analysis methods.Proteins' thermal transport, while being an important biophysical characteristic, lacks a definitive understanding regarding its interaction with protein structures, dynamics, and functional roles. Conformational fluctuations in folded proteins' structures are characterized by a highly inhomogeneous, anisotropic, and non-uniform thermal energy flow. upr signals inhibitor Equilibrium molecular dynamics simulations were used to analyze local thermal transport properties in proteins, employing a theoretical framework based on the autocorrelation function. We created a linear-homopolymer-like model, applying it to the villin headpiece subdomain (HP36), a small α-helical protein. Consequently, the model replicated the precise thermal conductivity of the protein, differing by less than 1%. A fascinating observation emerged from the site-selective analysis of thermal conductivity at the local residue level. The magnitude of thermal conductivity demonstrated a clear dependence on residue type, decreasing sequentially from charged to polar to hydrophobic residues. In conjunction with this, the residue-specific thermal transport property's sensitivity to the local density was also elaborated upon.This research paper makes use of the previously introduced Canonical Polyadic (CP)-Multiple Shift Block Inverse Iteration (MSBII) eigensolver, cited in [S]. The Journal of Chemical showcased the study conducted by D. Kallullathil and T. Carrington. Physically, I am experiencing a feeling of revitalization. Vibrational spectra were determined using 155, 234105 (2021) alongside a contraction tree algorithm. Employing the CP format, the CP-MSBII eigensolver operates. A linear relationship exists between the memory cost and the count of coordinates. A wavefunction expressed in CP format is inherently restricted to a sum-of-products (SOP) structure. An improvement in the accuracy of an SOP wavefunction is achievable by augmenting the number of terms and increasing its rank. Calculations performed in CP format, though conserving memory, exhibit a lengthy processing time when the necessary rank is substantial. By utilizing a contraction tree, the method's efficacy is improved as we dissect the entire problem into components. The rank requirement is modest for each of the sub-problems. In order to showcase the potency of the concepts, we measured the vibrational energy levels of acetonitrile (12-D) and ethylene oxide (15-D).Within nanoscale gas transport systems, fluid-surface interactions dominate the forces that control the evolution of the flow state. Carbon nanotubes with atomically smooth surfaces, typical of ideal nanoscale systems, have reduced characteristic lengths to such an extent that the non-equilibrium entrance region represents a major portion of the domain. This regime of operation sees the effective resistance, significantly amplified by the non-equilibrium entrance region, exceeding the limitations of classical effusion models. The intricacies of resistance within this operational framework are yet to be fully elucidated. Using a novel finite-difference solution, this paper develops a stochastic model of interfacial resistance, enabling the determination of the effective diffusion coefficient. Through this method, we model free-molecular gas flow within extensive nanotubes, revealing the possible existence of non-equilibrium effects in systems with dimensions compatible with current manufacturing techniques.