Rapid generation of antibody conformational ensembles using machine learning
- 1Genentech
Abstract
We introduce Cadense, a generative framework for antibody ensemble prediction that produces diverse Fv structures from sequence, informed by conformational dynamics. While a valuable resource, static structure prediction provides only a single snapshot of dynamic proteins, whose motions influence their binding and other properties. This is particularly relevant for antibodies, an important class of molecules for drug discovery, where flexible complementarity determining region (CDR) loops are difficult to model with static approaches. Cadense addresses this gap by generating ensembles that match conformational and biophysical property distributions from molecular dynamics (MD), while reducing the computational burden from days of simulation to seconds of generation. We show Cadense ensembles can recover both apo and holo conformations of the same antibody and serve as effective input to physics-based docking, capturing the conformational flexibility important to properly model antibody antigen interactions. Together, these results establish Cadense as a useful tool for antibody conformational ensemble generation with the promise to enable faster, dynamics-aware antibody design.
Related
- Pocket contrastive learning and conformation fusion for robust structure-conditioned 3D molecular generation under receptor conformational variation.
- DL3D: visual representation of conformational ensembles of domain-linker-domain proteins
- TopoFlow: Evolutionarily Conditioned Flow Matching for Protein Conformational Ensemble Generation
- Cryo-EM and X-ray structures of an anti-MHC-I monoclonal antibody/H2-Kb complex reveal the basis of conformational dependence and allelic specificity
- trFlow: ultrafast all-atom protein conformational ensemble generation via conditional flow matching
BibTeX
@misc{rao2026rapid,
title = {{Rapid generation of antibody conformational ensembles using machine learning}},
author = {P. V. Rao and D. Sommer and D. Davidson and R. G. Alberstein and J. Ludwiczak and J. Kleinhenz and B. Vani and E. Park and J. H. Lee and S. Izadi and J. R. Kiefer and R. Bonneau and S. P. Mahajan and A. M. Watkins and F. A. Dreyer},
howpublished = {bioRxiv},
year = {2026},
month = nov,
doi = {10.64898/2026.11.06.764889},
url = {https://www.biorxiv.org/content/10.64898/2026.11.06.764889v1},
note = {CC BY},
}
RIS
TY - UNPB
TI - Rapid generation of antibody conformational ensembles using machine learning
AU - P. V. Rao
AU - D. Sommer
AU - D. Davidson
AU - R. G. Alberstein
AU - J. Ludwiczak
AU - J. Kleinhenz
AU - B. Vani
AU - E. Park
AU - J. H. Lee
AU - S. Izadi
AU - J. R. Kiefer
AU - R. Bonneau
AU - S. P. Mahajan
AU - A. M. Watkins
AU - F. A. Dreyer
PY - 2026
DA - 2026/11/06
PB - bioRxiv
DO - 10.64898/2026.11.06.764889
UR - https://www.biorxiv.org/content/10.64898/2026.11.06.764889v1
AB - We introduce Cadense, a generative framework for antibody ensemble prediction that produces diverse Fv structures from sequence, informed by conformational dynamics. While a valuable resource, static structure prediction provides only a single snapshot of dynamic proteins, whose motions influence their binding and other properties. This is particularly relevant for antibodies, an important class of molecules for drug discovery, where flexible complementarity determining region (CDR) loops are difficult to model with static approaches. Cadense addresses this gap by generating ensembles that match conformational and biophysical property distributions from molecular dynamics (MD), while reducing the computational burden from days of simulation to seconds of generation. We show Cadense ensembles can recover both apo and holo conformations of the same antibody and serve as effective input to physics-based docking, capturing the conformational flexibility important to properly model antibody antigen interactions. Together, these results establish Cadense as a useful tool for antibody conformational ensemble generation with the promise to enable faster, dynamics-aware antibody design.
ER -