Saved in:
Bibliographic Details
Main Authors: Lhotka, Jirka, Probst, Daniel
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2502.10848
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866929717587738624
author Lhotka, Jirka
Probst, Daniel
author_facet Lhotka, Jirka
Probst, Daniel
contents Molecules have various computational representations, including numerical descriptors, strings, graphs, point clouds, and surfaces. Each representation method enables the application of various machine learning methodologies from linear regression to graph neural networks paired with large language models. To complement existing representations, we introduce the representation of molecules through vector-valued functions, or $n$-dimensional vector fields, that are parameterized by neural networks, which we denote molecular neural fields. Unlike surface representations, molecular neural fields capture external features and the hydrophobic core of macromolecules such as proteins. Compared to discrete graph or point representations, molecular neural fields are compact, resolution independent and inherently suited for interpolation in spatial and temporal dimensions. These properties inherited by molecular neural fields lend themselves to tasks including the generation of molecules based on their desired shape, structure, and composition, and the resolution-independent interpolation between molecular conformations in space and time. Here, we provide a framework and proofs-of-concept for molecular neural fields, namely, the parametrization and superresolution reconstruction of a protein-ligand complex using an auto-decoder architecture and the embedding of molecular volumes in latent space using an auto-encoder architecture.
format Preprint
id arxiv_https___arxiv_org_abs_2502_10848
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Implicit Neural Representations of Molecular Vector-Valued Functions
Lhotka, Jirka
Probst, Daniel
Machine Learning
Quantitative Methods
J.3; I.2.10
Molecules have various computational representations, including numerical descriptors, strings, graphs, point clouds, and surfaces. Each representation method enables the application of various machine learning methodologies from linear regression to graph neural networks paired with large language models. To complement existing representations, we introduce the representation of molecules through vector-valued functions, or $n$-dimensional vector fields, that are parameterized by neural networks, which we denote molecular neural fields. Unlike surface representations, molecular neural fields capture external features and the hydrophobic core of macromolecules such as proteins. Compared to discrete graph or point representations, molecular neural fields are compact, resolution independent and inherently suited for interpolation in spatial and temporal dimensions. These properties inherited by molecular neural fields lend themselves to tasks including the generation of molecules based on their desired shape, structure, and composition, and the resolution-independent interpolation between molecular conformations in space and time. Here, we provide a framework and proofs-of-concept for molecular neural fields, namely, the parametrization and superresolution reconstruction of a protein-ligand complex using an auto-decoder architecture and the embedding of molecular volumes in latent space using an auto-encoder architecture.
title Implicit Neural Representations of Molecular Vector-Valued Functions
topic Machine Learning
Quantitative Methods
J.3; I.2.10
url https://arxiv.org/abs/2502.10848