Intermolecular Non-Bonded Interactions from Machine Learning Datasets

Chen, Jia-An; Chao, Sheng D.

doi:10.3390/molecules28237900

Cited by 3 publications

(1 citation statement)

References 59 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our main purpose is to test a recently released ML algorithm called the CLIFF scheme [ 67 ]. In a previous study [ 68 ], we used a lower SAPT level of theory (SAPT0) to calculate the potential energy data and tested the ability of the CLIFF scheme to interpolate the datasets, with an emphasis on the possible overfitting problems. In this present paper, we would like to test the effect of changing the SAPT theory level on the predictive power and determine a minimum level of theory (SAPT2) for approaching the benchmark accuracy.…”

Section: Introductionmentioning

confidence: 99%

A Machine Learning Force Field for Bio-Macromolecular Modeling Based on Quantum Chemistry-Calculated Interaction Energy Datasets

Fan,

Chao

2024

Bioengineering

Self Cite

View full text Add to dashboard Cite

Accurate energy data from noncovalent interactions are essential for constructing force fields for molecular dynamics simulations of bio-macromolecular systems. There are two important practical issues in the construction of a reliable force field with the hope of balancing the desired chemical accuracy and working efficiency. One is to determine a suitable quantum chemistry level of theory for calculating interaction energies. The other is to use a suitable continuous energy function to model the quantum chemical energy data. For the first issue, we have recently calculated the intermolecular interaction energies using the SAPT0 level of theory, and we have systematically organized these energies into the ab initio SOFG-31 (homodimer) and SOFG-31-heterodimer datasets. In this work, we re-calculate these interaction energies by using the more advanced SAPT2 level of theory with a wider series of basis sets. Our purpose is to determine the SAPT level of theory proper for interaction energies with respect to the CCSD(T)/CBS benchmark chemical accuracy. Next, to utilize these energy datasets, we employ one of the well-developed machine learning techniques, called the CLIFF scheme, to construct a general-purpose force field for biomolecular dynamics simulations. Here we use the SOFG-31 dataset and the SOFG-31-heterodimer dataset as the training and test sets, respectively. Our results demonstrate that using the CLIFF scheme can reproduce a diverse range of dimeric interaction energy patterns with only a small training set. The overall errors for each SAPT energy component, as well as the SAPT total energy, are all well below the desired chemical accuracy of ~1 kcal/mol.

show abstract

Section: Introductionmentioning

confidence: 99%

A Machine Learning Force Field for Bio-Macromolecular Modeling Based on Quantum Chemistry-Calculated Interaction Energy Datasets

Fan,

Chao

2024

Bioengineering

Self Cite

View full text Add to dashboard Cite

show abstract

Unveiling the energetic complexity of noncovalent interactions in halogenated dimers

Liu,

2024

Int J of Quantum Chemistry

View full text Add to dashboard Cite

The understanding of noncovalent interactions is crucial in explaining critical phenomena such as self‐assembly, chemical reactivity, and crystallization. This work examines the energetic diversity of conformations and local minima for several halogenated dimers, represented as R‐X (R = H, F, CH3, CF3; X = Cl, Br, I). Thousands of configurations are randomly generated and refined through geometric optimizations to yield a diverse set of molecular conformers. Frequency calculations were performed for all optimized conformers to confirm that they are local minima. The noncovalent interactions in optimized dimers of halogen‐containing molecules were analyzed with atom in molecules (AIM) method and symmetry‐adapted perturbation theory (SAPT). Additionally, a protocol for generating machine learning models to recover accurate predictions of the physically meaningful SAPT energy components with minor computational cost is presented. These results deepen our understanding of the intricate energy balance and dedicated equilibrium of different noncovalent interactions in halogenated dimers.

show abstract

Simultaneous effects of temperature and backbone length on static and dynamic properties of high-density polyethylene-1-butene copolymer melt: Equilibrium molecular dynamics approach

Yazdanbakhsh,

Motlagh

2024

E-Polymers

View full text Add to dashboard Cite

Temperature and chain length play significant roles in determining the physical properties of polymer melts. In the current computational research, a molecular dynamics (MD) approach was implemented to describe the static and dynamic properties of (1) high-density polyethylene-1-butene with 120 beads in backbone (PE120) and (2) entangled high-density polyethylene-1-butene with 600 beads in the backbone (PE600). The transferable potentials for phase equilibria force fields were used for CH2 beads in a defined initial condition. First, the equilibrium phase of the designed systems was reported with total energy and density convergency at various initial temperatures (T 0 = 450, 470, and 490 K). Also, gyration radius (R g) and end-to-end distance (R) were calculated for the static behavior description of the two PEs. Zero-shear viscosity (η 0), mean square displacement, and diffusion coefficient (D) were estimated to define the dynamic behavior of PE120 and PE600 systems. MD outputs predicted that 10 ns is sufficient for equilibrium phase detection inside polymeric samples. After equilibrium phase detection, R g converged to 14.97 and 17.35 Å in PE120 and PE600, respectively (T 0 = 450 K). Furthermore, MD outputs show that temperature variation can considerably affect the time evolution of the system. Numerically, the η 0 of PE120 and PE600 converged to 49 and 168 cp at 450 K. These results of η 0 parameter as a function of temperature are an important output of MD simulations. The results predicted that η 0 decreases to 24 and 44 cp for PE120 and PE600 samples with an increase in temperature from 450 to 490 K. With the creation of the entanglements network, D reached the highest value of 2 × 10−9 m2·s−1 among the designed polymeric systems. The results are in good consistency with experimental reports. It is expected that the result of this study can be used in designing improved polymeric systems for real applications.

show abstract

Intermolecular Non-Bonded Interactions from Machine Learning Datasets

Cited by 3 publications

References 59 publications

A Machine Learning Force Field for Bio-Macromolecular Modeling Based on Quantum Chemistry-Calculated Interaction Energy Datasets

A Machine Learning Force Field for Bio-Macromolecular Modeling Based on Quantum Chemistry-Calculated Interaction Energy Datasets

Unveiling the energetic complexity of noncovalent interactions in halogenated dimers

Simultaneous effects of temperature and backbone length on static and dynamic properties of high-density polyethylene-1-butene copolymer melt: Equilibrium molecular dynamics approach

Contact Info

Product

Resources

About