Vibrational embedding theory.

We suggest a consistent framework for the embedding of reduced-space correlated vibrational wave functions in a potential of the remaining modes and generalize this concept to arbitrary many subspaces. We present an implementation of this framework for vibrational coupled-cluster theory and response treatments. For C=O stretches of small molecules, we show that the embedded treatment accelerates convergence for enlarging subsets. For the water dimer and trimer as well as a water wire in bacteriorhodopsin, we investigate different partitioning schemes for the embedding approach: In the local partitioning of the vibrations, the modes dominated by motions in the same spatial region are correlated, whereas in the energy-based partitioning, modes of similar fundamental frequencies are correlated. In most cases, we obtain better agreement with superset reference results for the local partitioning than for energy-based partitioning. This work represents an important step toward multi-level methodologies in vibrational-structure theory required for its application to sizable (bio-)molecular systems.

Fragmentation-Based Decomposition of a Metalloenzyme-Substrate Interaction: A Case Study for a Lytic Polysaccharide Monooxygenase.

We present a novel decomposition scheme for electronic interaction energies based on the flexible formulation of fragmentation schemes through fragment combination ranges (FCRs; J. Chem. Phys., 2021, 155, 164105). We devise a clear additive decomposition with contribution of nondisjoint fragments and correction terms for overlapping fragments and apply this scheme to the metalloenzyme-substrate complex of a lytic polysaccharide monooxygenase (LPMO) with an oligosaccharide. By this, we further illustrate the straightforward adaptability of the FCR-based schemes to novel systems. Our calculations suggest that the description of the electronic structure is a larger error source than the fragmentation scheme. In particular, we find a large impact of the basis set size on the interaction energies. Still, the introduction of three-body interaction terms in the fragmentation setup improves the agreement to the supermolecular reference. Yet, the qualitative results for the decomposition scheme with two-body terms only largely agree within the investigated electronic-structure approaches and basis sets, which are B97-3c, DFT (TPSS and B3LYP), and MP2 methods. The overlap contributions are found to be small, allowing analysis of the interaction energy into individual amino acid residues: We find a particularly strong interaction between the substrate and the LPMO copper active site.

A unified and flexible formulation of molecular fragmentation schemes.

We present a flexible formulation for energy-based molecular fragmentation schemes. This framework does not only incorporate the majority of existing fragmentation expansions but also allows for flexible formulation of novel schemes. We further illustrate its application in multi-level approaches and for electronic interaction energies. For the examples of small water clusters, a small protein, and protein-protein interaction energies, we show how this flexible setup can be exploited to generate a well-suited multi-level fragmentation expansion for the given case. With such a setup, we reproduce the electronic protein-protein interaction energy of ten different structures of a neurotensin and an extracellular loop of its receptor with a mean absolute deviation to the respective super-system calculations below 1 kJ/mol.