Interaction Energy of Large Molecules from Restrained Denominator MP2-F12

An Efficient RI-MP2 Algorithm for Distributed Many-GPU Architectures

Second-order Møller-Plesset perturbation theory (MP2) using the Resolution of the Identity approximation (RI-MP2) is a widely used method for computing molecular energies beyond the Hartree-Fock mean-field approximation. However, its high computational cost and lack of efficient algorithms for modern supercomputing architectures limit its applicability to large molecules. In this paper, we present the first distributed-memory many-GPU RI-MP2 algorithm explicitly designed to utilize hundreds of GPU accelerators for every step of the computation. Our novel algorithm achieves near-peak performance on GPU-based supercomputers through the development of a distributed memory algorithm for forming RI-MP2 intermediate tensors with zero internode communication, except for a single O ( N 2 ) asynchronous broadcast, and a distributed memory algorithm for the O ( N 5 ) energy reduction step, capable of sustaining near-peak performance on clusters with several hundred GPUs. Comparative analysis shows our implementation outperforms state-of-the-art quantum chemistry software by over 3.5 times in speed while achieving an 8-fold reduction in computational power consumption. Benchmarking on the Perlmutter supercomputer, our algorithm achieves 11.8 PFLOP/s (83% of peak performance) performing and the RI-MP2 energy calculation on a 314-water cluster with 7850 primary and 30,144 auxiliary basis functions in 4 min on 180 nodes and 720 A100 GPUs. This performance represents a substantial improvement over traditional CPU-based methods, demonstrating significant time-to-solution and power consumption benefits of leveraging modern GPU-accelerated computing environments for quantum chemistry calculations.

High-Performance Multi-GPU Analytic RI-MP2 Energy Gradients

This article presents a novel algorithm for the calculation of analytic energy gradients from second-order Møller-Plesset perturbation theory within the Resolution-of-the-Identity approximation (RI-MP2), which is designed to achieve high performance on clusters with multiple graphical processing units (GPUs). The algorithm uses GPUs for all major steps of the calculation, including integral generation, formation of all required intermediate tensors, solution of the Z-vector equation and gradient accumulation. The implementation in the EXtreme Scale Electronic Structure System (EXESS) software package includes a tailored, highly efficient, multistream scheduling system to hide CPU-GPU data transfer latencies and allows nodes with 8 A100 GPUs to operate at over 80% of theoretical peak floating-point performance. Comparative performance analysis shows a significant reduction in computational time relative to traditional multicore CPU-based methods, with our approach achieving up to a 95-fold speedup over the single-node performance of established software such as Q-Chem and ORCA. Additionally, we demonstrate that pairing our implementation with the molecular fragmentation framework in EXESS can drastically lower the computational scaling of RI-MP2 gradient calculations from quintic to subquadratic, enabling further substantial savings in runtime while retaining high numerical accuracy in the resulting gradients.

Coupled Cluster Benchmarking of Large Noncovalent Complexes in L7 and S12L as Well as the C60 Dimer, DNA-Ellipticine, and HIV-Indinavir

In this work, we report the benchmark binding energies of the seven complexes within the L7 data set, six host-guest complexes from the S12L data set, a C₆₀ dimer, the DNA-ellipticine intercalation complex, and the largest system of the study, the HIV-indinavir system, which contained 343 atoms or 139 heavy atoms. The high-quality values reported were obtained via a focal point method that relies on the canonical form of second-order Møller-Plesset theory and the domain-based local pair natural orbital scheme for the coupled cluster with single double and perturbative triple excitations [DLPNO-CCSD(T)] extrapolated to the complete basis set (CBS) limit. The results in this work not only corroborate but also improve upon some previous benchmark values for large noncovalent complexes albeit at a relatively steep cost. Although local CCSD(T) and the largely successful fixed-node diffusion Monte Carlo (FN-DMC) have been shown to generally agree for small- to medium-size systems, a discrepancy in their reported binding energy values arises for large complexes, where the magnitude of the disagreement is a definite cause for concern. For example, the largest deviation in the L7 data set was 2.8 kcal/mol (∼10%) on the low end in C3GC. Such a deviation only grows worse in the S12L set, which showed a difference of up to 10.4 kcal/mol (∼25%) by a conservative estimation in buckycatcher-C₆₀. The DNA-ellipticine complex also generated a disagreement of 4.4 kcal/mol (∼10%) between both state-of-the-art methods. The disagreement between local CCSD(T) and FN-DMC in large noncovalent complexes shows that it is urgently needed to have the canonical CCSD(T), the Monte Carlo CCSD(T), or the full configuration interaction quantum Monte Carlo approaches available to large systems on the hundred-atom scale to solve this dilemma. In addition, the performances of cheaper popular computational methods were assessed for the studied complexes with respect to DLPNO-CCSD(T)/CBS. r²SCAN-3c, B97M-V, and PBE0+D4 work well in large noncovalent complexes in this work, and GFN2-xTB performs well in π-π stacking complexes. B97M-V is the most reliable computationally efficient approach to predicting noncovalent interactions for large complexes, being the only one to have binding errors within the so-called 1 kcal/mol "chemical accuracy". The benchmark interaction energies of these host-guest complexes, molecular materials, and biological systems with electronic and medicinal implications provide crucial reference data for the improvement of current and future lower-cost methods.

Monte Carlo MP2-F12 for Noncovalent Interactions: The C60 Dimer

A scalable stochastic algorithm is presented that can evaluate explicitly correlated (F12) second-order many-body perturbation (MP2) energies of weak, noncovalent, intermolecular interactions. It first transforms the formulas of the MP2 and F12 energy differences into a short sum of high-dimensional integrals of Green's functions in real space and imaginary time. These integrals are then evaluated by the Monte Carlo method augmented by parallel execution, redundant-walker convergence acceleration, direct-sampling autocorrelation elimination, and control-variate error reduction. By sharing electron-pair walkers across the supermolecule and its subsystems spanned by the joint basis set, the statistical uncertainty is reduced by one to 2 orders of magnitude in the MP2 binding energy corrected for the basis-set incompleteness and superposition errors. The method predicts the MP2-F12/aug-cc-pVDZ binding energy of 19.1 ± 4.0 kcal mol^-1 for the C₆₀ dimer at the center distance of 9.748 Å.

Computational Study of the Addition of Methanethiol to 40+ Michael Acceptors as a Model for the Bioconjugation of Cysteines

A long series of Michael acceptors are studied computationally as potential alternatives to the maleimides that are used in most antibody-drug conjugates to link Cys of mAbs with cytotoxic drugs. The products of the reaction of methanethiol (CH3SH/MeSH, as a simple model of Cys) with N-methylated ethynesulfonamide, 2-ethynylpyridinium ion, propynamide, and methyl ethynephosphonamidate (that is, with HC≡C-EWG) are predicted by the M06-2X/6-311+G(d,p) method to be thermodynamically more stable, in relation to their precursors, than that of MeSH with N-methylmaleimide and, in general, with H2C═CH-EWG; calculations with AcCysOMe and tBuSH are also included. However, for the addition of the anion (MeS-), which is the reactive species, the order changes and N-methylated 2-vinylpyridinium ion, 2,3-butadienamide, and maleimide may give more easily the anionic adducts than several activated triple bonds; moreover, the calculated ΔG⧧ values increase following the order HC≡C-SO2NHMe, N-methylmaleimide, HC≡C-PO(OMe)NHMe, and HC≡C-CONHMe. In other words, MeS- is predicted to react more rapidly with maleimides than with ethynephosphonamidates and with propynamides, in agreement with the experimental results. New mechanistic details are disclosed regarding the advantageous use of some amides, especially of ethynesulfonamides, which, however, are more prone to double additions and exchange reactions.

Analytic gradients for the single-reference driven similarity renormalization group second-order perturbation theory

We derive and implement analytic energy gradients for the single-reference driven similarity renormalization group second-order perturbation theory (DSRG-PT2). The resulting equations possess an asymptotic scaling that is identical to that of the second-order Møller-Plesset perturbation theory (MP2), indicating that the exponential regularizer in the DSRG equations does not introduce formal difficulties in the gradient theory. We apply the DSRG-PT2 method to optimizing the geometries of 15 small molecules. The equilibrium bond lengths computed with DSRG-PT2 are found similar to those of MP2, yielding a mean absolute error of 0.0033 Å and a standard deviation of 0.0045 Å when compared with coupled cluster with singles, doubles, and perturbative triples.

Multireference Theories of Electron Correlation Based on the Driven Similarity Renormalization Group

The driven similarity renormalization group (DSRG) provides an alternative way to address the intruder state problem in quantum chemistry. In this review, we discuss recent developments of multireference methods based on the DSRG. We provide a pedagogical introduction to the DSRG and its various extensions and discuss its formal properties in great detail. In addition, we report several illustrative applications of the DSRG to molecular systems.

Communication: Almost error-free resolution-of-the-identity correlation methods by null space removal of the particle-hole interactions

We present a method to improve upon the resolution-of-the-identity (RI) for correlation methods. While RI is known to allow for drastic speedups, it relies on a cancellation of errors. Our method eliminates the errors introduced by RI which are known to be problematic for absolute energies. In this way, independence of the error compensation assumption for relative energies is also achieved. The proposed method is based on the idea of starting with an oversized RI basis and projecting out all of its unphysical parts. The approach can be easily implemented into existing RI codes and results in an overhead of about 30%, while effectively removing the RI error. In passing, this process alleviates the problem that for many frequently employed basis sets no optimized RI basis sets have been constructed. In this paper, the theory is presented and results are discussed exemplarily for the random phase approximation and Møller-Plesset perturbation theory.

Importance of the Electron Correlation and Dispersion Corrections in Calculations Involving Enamines, Hemiaminals, and Aminals. Comparison of B3LYP, M06-2X, MP2, and CCSD Results with Experimental Data

While B3LYP, M06-2X, and MP2 calculations predict the ΔG° values for exchange equilibria between enamines and ketones with similar acceptable accuracy, the M06-2X/6-311+G(d,p) and MP2/6-311+G(d,p) methods are required for enamine formation reactions (for example, for enamine 5a, arising from 3-methylbutanal and pyrrolidine). Stronger disagreement was observed when calculated energies of hemiaminals (N,O-acetals) and aminals (N,N-acetals) were compared with experimental equilibrium constants, which are reported here for the first time. Although it is known that the B3LYP method does not provide a good description of the London dispersion forces, while M06-2X and MP2 may overestimate them, it is shown here how large the gaps are and that at least single-point calculations at the CCSD(T)/6-31+G(d) level should be used for these reaction intermediates; CCSD(T)/6-31+G(d) and CCSD(T)/6-311+G(d,p) calculations afford ΔG° values in some cases quite close to MP2/6-311+G(d,p) while in others closer to M06-2X/6-311+G(d,p). The effect of solvents is similarly predicted by the SMD, CPCM, and IEFPCM approaches (with energy differences below 1 kcal/mol).