ProtComposer: Compositional Protein Structure Generation with 3D Ellipsoids

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Stark, Hannes, Jing, Bowen, Geffner, Tomas, Yim, Jason, Jaakkola, Tommi, Vahdat, Arash, Kreis, Karsten
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909528901025792
author Stark, Hannes
Jing, Bowen
Geffner, Tomas
Yim, Jason
Jaakkola, Tommi
Vahdat, Arash
Kreis, Karsten
author_facet Stark, Hannes
Jing, Bowen
Geffner, Tomas
Yim, Jason
Jaakkola, Tommi
Vahdat, Arash
Kreis, Karsten
contents We develop ProtComposer to generate protein structures conditioned on spatial protein layouts that are specified via a set of 3D ellipsoids capturing substructure shapes and semantics. At inference time, we condition on ellipsoids that are hand-constructed, extracted from existing proteins, or from a statistical model, with each option unlocking new capabilities. Hand-specifying ellipsoids enables users to control the location, size, orientation, secondary structure, and approximate shape of protein substructures. Conditioning on ellipsoids of existing proteins enables redesigning their substructure's connectivity or editing substructure properties. By conditioning on novel and diverse ellipsoid layouts from a simple statistical model, we improve protein generation with expanded Pareto frontiers between designability, novelty, and diversity. Further, this enables sampling designable proteins with a helix-fraction that matches PDB proteins, unlike existing generative models that commonly oversample conceptually simple helix bundles. Code is available at https://github.com/NVlabs/protcomposer.
format Preprint
id arxiv_https___arxiv_org_abs_2503_05025
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle ProtComposer: Compositional Protein Structure Generation with 3D Ellipsoids
Stark, Hannes
Jing, Bowen
Geffner, Tomas
Yim, Jason
Jaakkola, Tommi
Vahdat, Arash
Kreis, Karsten
Biomolecules
We develop ProtComposer to generate protein structures conditioned on spatial protein layouts that are specified via a set of 3D ellipsoids capturing substructure shapes and semantics. At inference time, we condition on ellipsoids that are hand-constructed, extracted from existing proteins, or from a statistical model, with each option unlocking new capabilities. Hand-specifying ellipsoids enables users to control the location, size, orientation, secondary structure, and approximate shape of protein substructures. Conditioning on ellipsoids of existing proteins enables redesigning their substructure's connectivity or editing substructure properties. By conditioning on novel and diverse ellipsoid layouts from a simple statistical model, we improve protein generation with expanded Pareto frontiers between designability, novelty, and diversity. Further, this enables sampling designable proteins with a helix-fraction that matches PDB proteins, unlike existing generative models that commonly oversample conceptually simple helix bundles. Code is available at https://github.com/NVlabs/protcomposer.
title ProtComposer: Compositional Protein Structure Generation with 3D Ellipsoids
topic Biomolecules
url https://arxiv.org/abs/2503.05025