arXiv subjects

Juno Nam

Publications and source records attributed to Juno Nam.

23 records · Page 2Linked to original sources

arXiv2024

Learning Collective Variables with Synthetic Data Augmentation through Physics-Inspired Geodesic Interpolation

In molecular dynamics simulations, rare events, such as protein folding, are typically studied using enhanced sampling techniques, most of which are based on the definition of a collective variable (CV) along which acceleration occurs. Obtaining an expressive CV is crucial, but often hindered by the lack of information about the particular event, e.g., the transition from unfolded to folded conformation. We propose a simulation-free data augmentation strategy using physics-inspired metrics to generate geodesic interpolations resembling protein folding transitions, thereby improving sampling efficiency without true transition state samples. This new data can be used to improve the accuracy of classifier-based methods. Alternatively, a regression-based learning scheme for CV models can be adopted by leveraging the interpolation progress parameter.

physics.chem-ph↗

arXiv2024

Understanding active learning of molecular docking and its applications

With the advancing capabilities of computational methodologies and resources, ultra-large-scale virtual screening via molecular docking has emerged as a prominent strategy for in silico hit discovery. Given the exhaustive nature of ultra-large-scale virtual screening, active learning methodologies have garnered attention as a means to mitigate computational cost through iterative small-scale docking and machine learning model training. While the efficacy of active learning methodologies has been empirically validated in extant literature, a critical investigation remains in how surrogate models can predict docking score without considering three-dimensional structural features, such as receptor conformation and binding poses. In this paper, we thus investigate how active learning methodologies effectively predict docking scores using only 2D structures and under what circumstances they may work particularly well through benchmark studies encompassing six receptor targets. Our findings suggest that surrogate models tend to memorize structural patterns prevalent in high docking scored compounds obtained during acquisition steps. Despite this tendency, surrogate models demonstrate utility in virtual screening, as exemplified in the identification of actives from DUD-E dataset and high docking-scored compounds from EnamineReal library, a significantly larger set than the initial screening pool. Our comprehensive analysis underscores the reliability and potential applicability of active learning methodologies in virtual screening campaigns.

cs.LG↗

arXiv2022

Enhanced Sampling for Free Energy Profiles with Post-Transition-State Bifurcations

We present a method to explore the free energy landscapes of chemical reactions with post-transition-state bifurcations using an enhanced sampling method based on well-tempered metadynamics. Obviating the need for computationally expensive DFT-level ab initio molecular dynamics simulations, we obtain accurate energetics by utilizing a free energy perturbation scheme and deep learning estimator for the single-point energies of substrate configurations. Using a pair of easily interpretable collective variables, we present a quantitative free energy surface that is compatible with harmonic transition state theory calculations and in which the bifurcations are clearly visible. We demonstrate our approach with the example of the SpnF-catalyzed Diels-Alder reaction, a cycloaddition reaction in which post-transition-state bifurcation leads to the [4+2] as well as the [6+4] cycloadduct. We obtain the free energy landscapes for different stereochemical reaction pathways and characterize the mechanistic continuum between relevant reaction channels without explicitly searching for the pertinent transition state structures.

physics.chem-ph↗

arXiv2020

Thermodynamic identities with sunray diagrams

One of the hurdles in teaching undergraduate thermodynamics is a plethora of complicated partial derivative identities. Students suffer from difficulties in deriving, justifying, or interpreting the identities, misconceptions about partial derivatives, and a lack of in-depth understanding of the meanings of identities. We propose a novel diagrammatic method for the calculus of differentials and partial derivatives called the "sunray diagram" that resolves the difficulties above. The sunray diagram technique relates a partial derivative with "arrow sliding," which enables an aesthetic and intuitive manipulation of partial derivative expressions in the form of successive arrow slidings. Furthermore, the sunray diagram is more than an ad hoc or abstract machinery but is based on the symplectic structure of thermodynamics; the sunray diagram admits a direct physical interpretation on the P-V (or T-S) plane. We elaborate on such physical semantics of the sunray diagram by taking Maxwell's approach to the geometry of thermodynamic structures--reinterpreted in terms of differential geometry--as a reference point. We anticipate that our discussion introduces the geometry of thermodynamics to learners and enriches the graphical pedagogy in physics education.

physics.ed-ph↗

arXiv2016

Linking the Neural Machine Translation and the Prediction of Organic Chemistry Reactions

Finding the main product of a chemical reaction is one of the important problems of organic chemistry. This paper describes a method of applying a neural machine translation model to the prediction of organic chemical reactions. In order to translate 'reactants and reagents' to 'products', a gated recurrent unit based sequence-to-sequence model and a parser to generate input tokens for model from reaction SMILES strings were built. Training sets are composed of reactions from the patent databases, and reactions manually generated applying the elementary reactions in an organic chemistry textbook of Wade. The trained models were tested by examples and problems in the textbook. The prediction process does not need manual encoding of rules (e.g., SMARTS transformations) to predict products, hence it only needs sufficient training reaction sets to learn new types of reactions.

cs.LG↗