Traditional protein force fields use one set of parameters for most of the 20 amino acids (AAs), allowing transferability of the parameters. However, a significant shortcoming is the difficulty to fit the Ramachandran plots of all AA residues simultaneously, affecting the accuracy of the force field. In this Feature Article, we report a new strategy for protein force field parametrization. Backbone and side-chain conformational distributions of all 20 AA residues obtained from protein coil library were used as the target data. The dihedral angle (torsion) potentials and some local nonbonded (1-4/1-5/1-6) interactions in OPLS-AA/L force field were modified such that the target data can be excellently reproduced by molecular dynamics simulations of dipeptides (blocked AAs) in explicit water, resulting in a new force field with AA-specific parameters, RSFF1. An efficient free energy decomposition approach was developed to separate the corrections on ϕ and ψ from the two-dimensional Ramachandran plots. RSFF1 is shown to reproduce the experimental NMR (3)J-coupling constants of AA dipeptides better than other force fields. It has a good balance between α-helical and β-sheet secondary structures. It can successfully fold a set of α-helix proteins (Trp-cage and Homeodomain) and β-hairpins (Trpzip-2, GB1 hairpin), which cannot be consistently stabilized by other state-of-the-art force fields. Interestingly, the RSFF1 force field systematically overestimates the melting temperature (and the stability of native state) of these peptides/proteins. It has a potential application in the simulation of protein folding and protein structure refinement.

Recently, we developed a residue-specific force field (RSFF1) based on conformational free-energy distributions of the 20 amino acid residues from a protein coil library. Most parameters in RSFF1 were adopted from the OPLS-AA/L force field, but some van der Waals and torsional parameters that effectively affect local conformational preferences were introduced specifically for individual residues to fit the coil library distributions. Here a similar strategy has been applied to modify the Amber ff99SB force field, and a new force field named RSFF2 is developed. It can successfully fold α-helical structures such as polyalanine peptides, Trp-cage miniprotein, and villin headpiece subdomain and β-sheet structures such as Trpzip-2, GB1 β-hairpins, and the WW domain, simultaneously. The properties of various popular force fields in balancing between α-helix and β-sheet are analyzed based on their descriptions of local conformational features of various residues, and the analysis reveals the importance of accurate local free-energy distributions. Unlike the RSFF1, which overestimates the stability of both α-helix and β-sheet, RSFF2 gives melting curves of α-helical peptides and Trp-cage in good agreement with experimental data. Fitting to the two-state model, RSFF2 gives folding enthalpies and entropies in reasonably good agreement with available experimental results.

Chalcopyrite compound CuGaTe2 is the focus of much research interest due to its high power factor. However, its high intrinsic lattice thermal conductivity seriously impedes the promotion of its thermoelectric performance. Here, it is shown that through alloying of isoelectronic elements In and Ag in CuGaTe2, a quinary alloy compound system Cu1−xAgxGa0.4In0.6Te2 (0 ≤ x ≤ 0.4) with complex nanosized strain domain structure is prepared. Due to strong phonon scattering mainly by this domain structure, thermal conductivity (at 300 K) drops from 6.1 W m−1 K−1 for the host compound to 1.5 W m−1 K−1 for the sample with x = 0.4. As a result, the optimized chalcopyrite sample Cu0.7Ag0.3Ga0.4In0.6Te2 presents an outstanding performance, with record‐high figure of merit (ZT) reaching 1.64 (at 873 K) and average ZT reaching 0.73 (between ≈300 and 873 K), which are ≈37 and ≈35% larger than the corresponding values for pristine CuGaTe2, respectively, demonstrating that such domain structure arising from isoelectronic multielement alloying in chalcopyrite compound can effectively suppress its thermal conductivity and elevate its thermoelectric performance remarkably.

No abstract

No abstract