Perspective: Advances, Challenges, and Insight for Predictive Coarse-Grained Models

Abstract

By averaging over atomic details, coarse-grained (CG) models provide profound computational and conceptual advantages for studying soft materials. In particular, bottom-up approaches develop CG models based upon information obtained from atomically detailed models. At least in principle, a bottom-up model can reproduce all the properties of an atomically detailed model that are observable at the resolution of the CG model. Historically, bottom-up approaches have accurately modeled the structure of liquids, polymers, and other amorphous soft materials, but have provided lower structural fidelity for more complex biomolecular systems. Moreover, they have also been plagued by unpredictable transferability and a poor description of thermodynamic properties. Fortunately, recent studies have reported dramatic advances in addressing these prior limitations. This Perspective reviews this remarkable progress, while focusing on its foundation in the basic theory of coarse-graining. In particular, we describe recent insights and advances for treating the CG mapping, for modeling many-body interactions, for addressing the state-point dependence of effective potentials, and even for reproducing atomic observables that are beyond the resolution of the CG model. We also outline outstanding challenges and promising directions in the field. We anticipate that the synthesis of rigorous theory and modern computational tools will result in practical bottom-up methods that not only are accurate and transferable but also provide predictive insight for complex systems.

Improved Protein Model in SPICA Force Field

“…Overall, SPICA FF ver2 is applicable with high accuracy for simulating extensive protein systems. Specifically, our new strategy of using secondary structure-dependent LJ parameters for protein BB segments overcomes known limitations of existing “pragmatic” FF without the need of specific ad hoc rescaling or parametrization strategies. ,, Our work suggests that there is still ample room for improvement toward the accuracy of chemically transferable pragmatic CG force fields using chemistry-driven intuition.…”

Pragmatic Coarse-Graining of Proteins: Models and Applications

“…The “coarser” pragmatic models will represent the backbone and side chains using a single bead each (e.g., UNRES), while “finer” models will use several beads for either moiety (e.g., SIRAH). Unlike mapping strategies often employed by other more abstract CG protein models, ,− the consistent and chemical-based manner by which “pragmatic” models represent their protein chain facilitates both capturing specific chemical information (e.g., specific interactions between side chains) and interpreting simulation results. The pragmatic nature of these models is also reflected in their implementation and parametrization.…”

“…For a review of the advances in CG protein modeling up to 2013, see the review by Baaden and Marrink . For a review of the major CG protein models and their applications pre-2017, see the comprehensive review by Kmiecik et al Other reviews have also covered more basic theoretical notions of coarse graining theory, addressing some of the concepts touched upon during this introduction in more detail. − …”

“…where each residue is represented by at least two beads (except for small amino acid residues, such as glycine), have a mostly physical-based implementation, and for which the parametrization was at least in part based on a top-down experimental approach. This kind of CG protein model has been defined as "pragmatic" by William Noid in a recent review, 44 a terminology which we will use and expand upon in the following section. We will cover in detail the main specifications and features of four of the major CG models within this scope�Martini, SIRAH, SPICA, and UNRES�as well as their most relevant applications since 2017.…”

Coarse-Grained Modeling Using Neural Networks Trained on Structural Data

“…Also, the relative entropy minimization method has been applied to train ANN for CG models . An overview of some other approaches that use ML in coarse-graining is given in reviews. , Furhermore, perspectives of coarse-grained ANN force fields are discussed in recent papers. , …”