---
DOAJ_listed: '1'
OA_place: publisher
OA_type: gold
PlanS_conform: '1'
_id: '20100'
abstract:
- lang: eng
text: A key step in protein structure prediction involves the detection of co-evolving
pairs of residues, a signal for spatial proximity. This information is gleaned
from multiple sequence alignment and underscores Alphafold’s structure prediction
for almost every known protein. A simple means to create proteins beyond those
found in nature, is by unnaturally fusing together two known proteins or protein
parts. Here we demonstrate that structured peptides are predicted with significantly
reduced accuracy when added to the terminal ends of scaffold proteins. Appending
the multiple sequence alignment for the individual peptide tags to that of the
scaffold protein often restores prediction accuracy. This work suggests that this
windowed multiple sequence alignment approach can be a useful tool for predicting
the structure of fused, chimeric proteins.
acknowledgement: AM acknowledges the financial support of the Helmsley Fellowships
Program for Sustainability and Health. AMB is supported by the Schmidt Chair in
Artificial Intelligence.
article_processing_charge: Yes
article_type: original
author:
- first_name: Sanketh
full_name: Vedula, Sanketh
id: 94f2fe44-70fa-11f0-b76b-92922c09452b
last_name: Vedula
- first_name: Alexander
full_name: Bronstein, Alexander
id: 58f3726e-7cba-11ef-ad8b-e6e8cb3904e6
last_name: Bronstein
orcid: 0000-0001-9699-8730
- first_name: Ailie
full_name: Marx, Ailie
last_name: Marx
citation:
ama: Vedula S, Bronstein AM, Marx A. Improving prediction accuracy in chimeric proteins
with windowed multiple sequence alignment. Computational and Structural Biotechnology
Journal. 2025;27:3292-3298. doi:10.1016/j.csbj.2025.07.039
apa: Vedula, S., Bronstein, A. M., & Marx, A. (2025). Improving prediction accuracy
in chimeric proteins with windowed multiple sequence alignment. Computational
and Structural Biotechnology Journal. Elsevier. https://doi.org/10.1016/j.csbj.2025.07.039
chicago: Vedula, Sanketh, Alex M. Bronstein, and Ailie Marx. “Improving Prediction
Accuracy in Chimeric Proteins with Windowed Multiple Sequence Alignment.” Computational
and Structural Biotechnology Journal. Elsevier, 2025. https://doi.org/10.1016/j.csbj.2025.07.039.
ieee: S. Vedula, A. M. Bronstein, and A. Marx, “Improving prediction accuracy in
chimeric proteins with windowed multiple sequence alignment,” Computational
and Structural Biotechnology Journal, vol. 27. Elsevier, pp. 3292–3298, 2025.
ista: Vedula S, Bronstein AM, Marx A. 2025. Improving prediction accuracy in chimeric
proteins with windowed multiple sequence alignment. Computational and Structural
Biotechnology Journal. 27, 3292–3298.
mla: Vedula, Sanketh, et al. “Improving Prediction Accuracy in Chimeric Proteins
with Windowed Multiple Sequence Alignment.” Computational and Structural Biotechnology
Journal, vol. 27, Elsevier, 2025, pp. 3292–98, doi:10.1016/j.csbj.2025.07.039.
short: S. Vedula, A.M. Bronstein, A. Marx, Computational and Structural Biotechnology
Journal 27 (2025) 3292–3298.
date_created: 2025-08-03T22:01:31Z
date_published: 2025-06-27T00:00:00Z
date_updated: 2025-11-27T14:09:59Z
day: '27'
ddc:
- '000'
- '570'
department:
- _id: AlBr
doi: 10.1016/j.csbj.2025.07.039
external_id:
isi:
- '001583543100001'
file:
- access_level: open_access
checksum: 78d01f30fc1dc11dd2bd1d7bb7ac8a62
content_type: application/pdf
creator: dernst
date_created: 2025-08-04T06:25:23Z
date_updated: 2025-08-04T06:25:23Z
file_id: '20104'
file_name: 2025_CompStrucBiotechJour_Vedula.pdf
file_size: 6609770
relation: main_file
success: 1
file_date_updated: 2025-08-04T06:25:23Z
has_accepted_license: '1'
intvolume: ' 27'
isi: 1
language:
- iso: eng
month: '06'
oa: 1
oa_version: Published Version
page: 3292-3298
publication: Computational and Structural Biotechnology Journal
publication_identifier:
eissn:
- 2001-0370
publication_status: published
publisher: Elsevier
quality_controlled: '1'
related_material:
link:
- relation: software
url: https://github.com/sankethvedula/AFChimera
record:
- id: '20103'
relation: software
status: public
scopus_import: '1'
status: public
title: Improving prediction accuracy in chimeric proteins with windowed multiple sequence
alignment
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
short: CC BY (4.0)
type: journal_article
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 27
year: '2025'
...
---
OA_place: repository
_id: '20103'
abstract:
- lang: eng
text: Official implementation, windowed MSAs, and the predictions as reported in
the manuscript titled "Improving Prediction Accuracy in Chimeric Proteins with
Windowed Multiple Sequence Alignment". (2025-06-27)
article_processing_charge: No
author:
- first_name: Sanketh
full_name: Vedula, Sanketh
id: 94f2fe44-70fa-11f0-b76b-92922c09452b
last_name: Vedula
- first_name: Alexander
full_name: Bronstein, Alexander
id: 58f3726e-7cba-11ef-ad8b-e6e8cb3904e6
last_name: Bronstein
orcid: 0000-0001-9699-8730
- first_name: Ailie
full_name: Marx, Ailie
last_name: Marx
citation:
ama: 'Vedula S, Bronstein AM, Marx A. Replication Data for: “Improving Prediction
Accuracy in Chimeric Proteins with Windowed Multiple Sequence Alignment.” 2025.
doi:10.7910/DVN/DYEBVM'
apa: 'Vedula, S., Bronstein, A. M., & Marx, A. (2025). Replication Data for:
“Improving Prediction Accuracy in Chimeric Proteins with Windowed Multiple Sequence
Alignment.” Harvard Dataverse. https://doi.org/10.7910/DVN/DYEBVM'
chicago: 'Vedula, Sanketh, Alex M. Bronstein, and Ailie Marx. “Replication Data
for: ‘Improving Prediction Accuracy in Chimeric Proteins with Windowed Multiple
Sequence Alignment.’” Harvard Dataverse, 2025. https://doi.org/10.7910/DVN/DYEBVM.'
ieee: 'S. Vedula, A. M. Bronstein, and A. Marx, “Replication Data for: ‘Improving
Prediction Accuracy in Chimeric Proteins with Windowed Multiple Sequence Alignment.’”
Harvard Dataverse, 2025.'
ista: 'Vedula S, Bronstein AM, Marx A. 2025. Replication Data for: ‘Improving Prediction
Accuracy in Chimeric Proteins with Windowed Multiple Sequence Alignment’, Harvard
Dataverse, 10.7910/DVN/DYEBVM.'
mla: 'Vedula, Sanketh, et al. Replication Data for: “Improving Prediction Accuracy
in Chimeric Proteins with Windowed Multiple Sequence Alignment.” Harvard Dataverse,
2025, doi:10.7910/DVN/DYEBVM.'
short: S. Vedula, A.M. Bronstein, A. Marx, (2025).
date_created: 2025-08-04T06:18:55Z
date_published: 2025-06-27T00:00:00Z
date_updated: 2025-11-27T14:09:58Z
day: '27'
ddc:
- '000'
department:
- _id: AlBr
doi: 10.7910/DVN/DYEBVM
has_accepted_license: '1'
license: https://creativecommons.org/publicdomain/zero/1.0/
main_file_link:
- open_access: '1'
url: https://doi.org/10.7910/DVN/DYEBVM
month: '06'
oa: 1
oa_version: Published Version
publisher: Harvard Dataverse
related_material:
record:
- id: '20100'
relation: used_for_analysis_in
status: public
status: public
title: 'Replication Data for: "Improving Prediction Accuracy in Chimeric Proteins
with Windowed Multiple Sequence Alignment"'
tmp:
image: /images/cc_0.png
legal_code_url: https://creativecommons.org/publicdomain/zero/1.0/legalcode
name: Creative Commons Public Domain Dedication (CC0 1.0)
short: CC0 (1.0)
type: research_data_reference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
year: '2025'
...
---
OA_place: publisher
OA_type: hybrid
_id: '20707'
abstract:
- lang: eng
text: 'Understanding physiological responses during running is critical for performance
optimization, tailored training prescriptions, and athlete health management.
We introduce a comprehensive framework—what we believe to be the first capable
of predicting instantaneous oxygen consumption (VO2) trajectories exclusively
from consumer-grade wearable data. Our approach employs two complementary physiological
models: (1) accurate modeling of heart rate (HR) dynamics via a physiologically
constrained ordinary differential equation (ODE) and neural Kalman filter, trained
on over 3 million HR observations, achieving 1-second interval predictions with
mean absolute errors as low as 2.81 bpm (correlation 0.87); and (2) leveraging
the principles of precise HR modeling, a novel VO2 prediction architecture requiring
only the initial second of VO2 data for calibration, enabling robust, sequence-to-sequence
metabolic demand estimation. Despite relying solely on smartwatch and chest-strap
data, our method achieves mean absolute percentage errors of approximately 13%,
effectively capturing rapid physiological transitions and steady-state conditions
across diverse running intensities. Our synchronized dataset, complemented by
blood lactate measurements, further lays the foundation for future noninvasive
metabolic zone identification. By embedding physiological constraints within modern
machine learning, this framework democratizes advanced metabolic monitoring, bridging
laboratory-grade accuracy and everyday accessibility, thus empowering both elite
athletes and recreational fitness enthusiasts.'
article_processing_charge: No
arxiv: 1
author:
- first_name: Barak
full_name: Gahtan, Barak
last_name: Gahtan
- first_name: Sanketh
full_name: Vedula, Sanketh
last_name: Vedula
- first_name: Gil
full_name: Samuelly Leichtag, Gil
last_name: Samuelly Leichtag
- first_name: Einat
full_name: Kodesh, Einat
last_name: Kodesh
- first_name: Alexander
full_name: Bronstein, Alexander
id: 58f3726e-7cba-11ef-ad8b-e6e8cb3904e6
last_name: Bronstein
orcid: 0000-0001-9699-8730
citation:
ama: 'Gahtan B, Vedula S, Samuelly Leichtag G, Kodesh E, Bronstein AM. From lab
to wrist: Bridging metabolic monitoring and consumer wearables for heart rate
and oxygen consumption modeling. In: Proceedings of the 27th International
Conference on Multimodal Interaction. Association for Computing Machinery;
2025:60-77. doi:10.1145/3716553.3750815'
apa: 'Gahtan, B., Vedula, S., Samuelly Leichtag, G., Kodesh, E., & Bronstein,
A. M. (2025). From lab to wrist: Bridging metabolic monitoring and consumer wearables
for heart rate and oxygen consumption modeling. In Proceedings of the 27th
International Conference on Multimodal Interaction (pp. 60–77). Canberra,
Australia: Association for Computing Machinery. https://doi.org/10.1145/3716553.3750815'
chicago: 'Gahtan, Barak, Sanketh Vedula, Gil Samuelly Leichtag, Einat Kodesh, and
Alex M. Bronstein. “From Lab to Wrist: Bridging Metabolic Monitoring and Consumer
Wearables for Heart Rate and Oxygen Consumption Modeling.” In Proceedings of
the 27th International Conference on Multimodal Interaction, 60–77. Association
for Computing Machinery, 2025. https://doi.org/10.1145/3716553.3750815.'
ieee: 'B. Gahtan, S. Vedula, G. Samuelly Leichtag, E. Kodesh, and A. M. Bronstein,
“From lab to wrist: Bridging metabolic monitoring and consumer wearables for heart
rate and oxygen consumption modeling,” in Proceedings of the 27th International
Conference on Multimodal Interaction, Canberra, Australia, 2025, pp. 60–77.'
ista: 'Gahtan B, Vedula S, Samuelly Leichtag G, Kodesh E, Bronstein AM. 2025. From
lab to wrist: Bridging metabolic monitoring and consumer wearables for heart rate
and oxygen consumption modeling. Proceedings of the 27th International Conference
on Multimodal Interaction. ICMI: International Conference on Multimodal Interaction,
60–77.'
mla: 'Gahtan, Barak, et al. “From Lab to Wrist: Bridging Metabolic Monitoring and
Consumer Wearables for Heart Rate and Oxygen Consumption Modeling.” Proceedings
of the 27th International Conference on Multimodal Interaction, Association
for Computing Machinery, 2025, pp. 60–77, doi:10.1145/3716553.3750815.'
short: B. Gahtan, S. Vedula, G. Samuelly Leichtag, E. Kodesh, A.M. Bronstein, in:,
Proceedings of the 27th International Conference on Multimodal Interaction, Association
for Computing Machinery, 2025, pp. 60–77.
conference:
end_date: 2025-10-17
location: Canberra, Australia
name: 'ICMI: International Conference on Multimodal Interaction'
start_date: 2025-10-13
corr_author: '1'
date_created: 2025-11-30T23:02:08Z
date_published: 2025-10-12T00:00:00Z
date_updated: 2025-12-01T07:22:09Z
day: '12'
ddc:
- '000'
department:
- _id: AlBr
doi: 10.1145/3716553.3750815
external_id:
arxiv:
- '2505.00101'
file:
- access_level: open_access
checksum: f793472a71d27012244567b499a4967f
content_type: application/pdf
creator: dernst
date_created: 2025-12-01T07:19:06Z
date_updated: 2025-12-01T07:19:06Z
file_id: '20713'
file_name: 2025_ICMI_Gahtan.pdf
file_size: 3045062
relation: main_file
success: 1
file_date_updated: 2025-12-01T07:19:06Z
has_accepted_license: '1'
language:
- iso: eng
month: '10'
oa: 1
oa_version: Published Version
page: 60-77
publication: Proceedings of the 27th International Conference on Multimodal Interaction
publication_identifier:
isbn:
- '9798400714993'
publication_status: published
publisher: Association for Computing Machinery
quality_controlled: '1'
scopus_import: '1'
status: public
title: 'From lab to wrist: Bridging metabolic monitoring and consumer wearables for
heart rate and oxygen consumption modeling'
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
short: CC BY (4.0)
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
year: '2025'
...
---
OA_place: publisher
OA_type: hybrid
_id: '19595'
abstract:
- lang: eng
text: We investigate the locality of magnetic response in polycyclic aromatic molecules
using a novel deep-learning approach. Our method employs graph neural networks
(GNNs) with a graph-of-rings representation to predict nucleus independent chemical
shifts (NICS) in the space around the molecule. We train a series of models, each
time reducing the size of the largest molecules used in training. The accuracy
of prediction remains high (MAE < 0.5 ppm), even when training the model only
on molecules with up to four rings, thus providing strong evidence for the locality
of magnetic response. To overcome the known problem of generalization of GNNs,
we implement a k-hop expansion strategy and succeed in achieving accurate predictions
for molecules with up to 15 rings (almost 4 times the size of the largest training
example). Our findings have implications for understanding the magnetic response
in complex molecules and demonstrate a promising approach to overcoming GNN scalability
limitations. Furthermore, the trained models enable rapid characterization, without
the need for more expensive DFT calculations.
acknowledgement: The authors express their gratitude to Professor Dr. Peter Chen for
his continued support. The authors acknowledge the Branco Weiss Fellowship for supporting
this research as part of a Society in Science grant and the Israel Science Foundation
for financial support (Grant No. 1745/23 to R.G.-P.). R.G.-P. is a Branco Weiss
Fellow, a Horev Fellow, and an Alon Scholarship recipient. A.M.B. was supported
by the ERC StG EARS and the Israeli Science Foundation.
article_number: '144101'
article_processing_charge: Yes (in subscription journal)
article_type: original
author:
- first_name: Yair
full_name: Davidson, Yair
last_name: Davidson
- first_name: Aviad
full_name: Philipp, Aviad
last_name: Philipp
- first_name: Sabyasachi
full_name: Chakraborty, Sabyasachi
last_name: Chakraborty
- first_name: Alexander
full_name: Bronstein, Alexander
id: 58f3726e-7cba-11ef-ad8b-e6e8cb3904e6
last_name: Bronstein
orcid: 0000-0001-9699-8730
- first_name: Renana
full_name: Gershoni-Poranne, Renana
last_name: Gershoni-Poranne
citation:
ama: Davidson Y, Philipp A, Chakraborty S, Bronstein AM, Gershoni-Poranne R. How
local is “local”? Deep learning reveals locality of the induced magnetic field
of polycyclic aromatic hydrocarbons. Journal of Chemical Physics. 2025;162(14).
doi:10.1063/5.0257558
apa: Davidson, Y., Philipp, A., Chakraborty, S., Bronstein, A. M., & Gershoni-Poranne,
R. (2025). How local is “local”? Deep learning reveals locality of the induced
magnetic field of polycyclic aromatic hydrocarbons. Journal of Chemical Physics.
AIP Publishing. https://doi.org/10.1063/5.0257558
chicago: Davidson, Yair, Aviad Philipp, Sabyasachi Chakraborty, Alex M. Bronstein,
and Renana Gershoni-Poranne. “How Local Is ‘Local’? Deep Learning Reveals Locality
of the Induced Magnetic Field of Polycyclic Aromatic Hydrocarbons.” Journal
of Chemical Physics. AIP Publishing, 2025. https://doi.org/10.1063/5.0257558.
ieee: Y. Davidson, A. Philipp, S. Chakraborty, A. M. Bronstein, and R. Gershoni-Poranne,
“How local is ‘local’? Deep learning reveals locality of the induced magnetic
field of polycyclic aromatic hydrocarbons,” Journal of Chemical Physics,
vol. 162, no. 14. AIP Publishing, 2025.
ista: Davidson Y, Philipp A, Chakraborty S, Bronstein AM, Gershoni-Poranne R. 2025.
How local is “local”? Deep learning reveals locality of the induced magnetic field
of polycyclic aromatic hydrocarbons. Journal of Chemical Physics. 162(14), 144101.
mla: Davidson, Yair, et al. “How Local Is ‘Local’? Deep Learning Reveals Locality
of the Induced Magnetic Field of Polycyclic Aromatic Hydrocarbons.” Journal
of Chemical Physics, vol. 162, no. 14, 144101, AIP Publishing, 2025, doi:10.1063/5.0257558.
short: Y. Davidson, A. Philipp, S. Chakraborty, A.M. Bronstein, R. Gershoni-Poranne,
Journal of Chemical Physics 162 (2025).
corr_author: '1'
date_created: 2025-04-20T22:01:28Z
date_published: 2025-04-14T00:00:00Z
date_updated: 2025-09-30T12:06:51Z
day: '14'
ddc:
- '000'
department:
- _id: AlBr
doi: 10.1063/5.0257558
external_id:
isi:
- '001466311300030'
pmid:
- '40197568'
file:
- access_level: open_access
checksum: 20a31a4c506b52de863bab7d3ff989ef
content_type: application/pdf
creator: dernst
date_created: 2025-04-22T09:27:43Z
date_updated: 2025-04-22T09:27:43Z
file_id: '19606'
file_name: 2025_JourChemicalPhysics_Davidson.pdf
file_size: 7812182
relation: main_file
success: 1
file_date_updated: 2025-04-22T09:27:43Z
has_accepted_license: '1'
intvolume: ' 162'
isi: 1
issue: '14'
language:
- iso: eng
month: '04'
oa: 1
oa_version: Published Version
pmid: 1
project:
- _id: 92f4a086-16d5-11f0-9cad-c929f5c58b0c
grant_number: '863839'
name: Acoustics-based drone navigation and interaction
publication: Journal of Chemical Physics
publication_identifier:
eissn:
- 1089-7690
issn:
- 0021-9606
publication_status: published
publisher: AIP Publishing
quality_controlled: '1'
related_material:
link:
- relation: software
url: https://gitlab.com/porannegroup/magnetic_locality
scopus_import: '1'
status: public
title: How local is “local”? Deep learning reveals locality of the induced magnetic
field of polycyclic aromatic hydrocarbons
tmp:
image: /images/cc_by_nc.png
legal_code_url: https://creativecommons.org/licenses/by-nc/4.0/legalcode
name: Creative Commons Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)
short: CC BY-NC (4.0)
type: journal_article
user_id: 317138e5-6ab7-11ef-aa6d-ffef3953e345
volume: 162
year: '2025'
...
---
DOAJ_listed: '1'
OA_place: publisher
OA_type: gold
PlanS_conform: '1'
_id: '21246'
abstract:
- lang: eng
text: Stellar astrophysics relies on diverse observational modalities—primarily
photometric light curves and spectroscopic data—from which fundamental stellar
properties are inferred. While machine learning (ML) has advanced analysis within
individual modalities, the complementary information encoded across modalities
remains largely underexploited. We present the dual embedding for stellar astronomy
(DESA) model, a novel multimodal foundation model that integrates light curves
and spectra to learn a unified, physically meaningful latent space for stars.
DESA first trains separate modality-specific encoders using a hybrid supervised/self-supervised
scheme, and then aligns them through DualFormer, a transformer-based cross-modal
integration module tailored for astrophysical data. DualFormer combines cross-
and self-attention, a novel dual-projection alignment loss, and a projection-space
eigendecomposition that yields physically structured embeddings. We demonstrate
that DESA significantly outperforms leading unimodal and self-supervised baselines
across a range of tasks. In zero- and few-shot settings, DESA’s learned representations
recover stellar color–magnitude and Hertzsprung–Russell diagrams with high fidelity
(R2 = 0.92 for photometric regressions). In full fine-tuning, DESA achieves state-of-the-art
accuracy for binary star detection (AUC = 0.99, AP = 1.00) and stellar age prediction
(RMSE = 0.94 Gyr). As a compelling case, DESA naturally separates synchronized
binaries from young stars—two populations with nearly identical light curves—purely
from their embedded positions in UMAP space, without requiring external kinematic
or luminosity information. DESA thus offers a powerful new framework for multimodal,
data-driven stellar population analysis, enabling both accurate prediction and
novel discovery.
acknowledgement: This research was partially supported by the Israeli Science Foundation
grant 1834/24.
article_number: '110'
article_processing_charge: Yes
article_type: original
arxiv: 1
author:
- first_name: Ilay
full_name: Kamai, Ilay
last_name: Kamai
- first_name: Alexander
full_name: Bronstein, Alexander
id: 58f3726e-7cba-11ef-ad8b-e6e8cb3904e6
last_name: Bronstein
orcid: 0000-0001-9699-8730
- first_name: Hagai B.
full_name: Perets, Hagai B.
last_name: Perets
citation:
ama: Kamai I, Bronstein AM, Perets HB. Machine Learning inference of stellar properties
using integrated photometric and spectroscopic data. The Astrophysical Journal.
2025;994. doi:10.3847/1538-4357/ae0cbc
apa: Kamai, I., Bronstein, A. M., & Perets, H. B. (2025). Machine Learning inference
of stellar properties using integrated photometric and spectroscopic data. The
Astrophysical Journal. IOP Publishing. https://doi.org/10.3847/1538-4357/ae0cbc
chicago: Kamai, Ilay, Alex M. Bronstein, and Hagai B. Perets. “Machine Learning
Inference of Stellar Properties Using Integrated Photometric and Spectroscopic
Data.” The Astrophysical Journal. IOP Publishing, 2025. https://doi.org/10.3847/1538-4357/ae0cbc.
ieee: I. Kamai, A. M. Bronstein, and H. B. Perets, “Machine Learning inference of
stellar properties using integrated photometric and spectroscopic data,” The
Astrophysical Journal, vol. 994. IOP Publishing, 2025.
ista: Kamai I, Bronstein AM, Perets HB. 2025. Machine Learning inference of stellar
properties using integrated photometric and spectroscopic data. The Astrophysical
Journal. 994, 110.
mla: Kamai, Ilay, et al. “Machine Learning Inference of Stellar Properties Using
Integrated Photometric and Spectroscopic Data.” The Astrophysical Journal,
vol. 994, 110, IOP Publishing, 2025, doi:10.3847/1538-4357/ae0cbc.
short: I. Kamai, A.M. Bronstein, H.B. Perets, The Astrophysical Journal 994 (2025).
date_created: 2026-02-16T15:35:29Z
date_published: 2025-11-19T00:00:00Z
date_updated: 2026-02-17T12:33:19Z
day: '19'
ddc:
- '520'
- '000'
department:
- _id: AlBr
doi: 10.3847/1538-4357/ae0cbc
external_id:
arxiv:
- '2507.10666'
file:
- access_level: open_access
checksum: 255ffd6d664e6c2d1cffbaced650bd10
content_type: application/pdf
creator: dernst
date_created: 2026-02-17T12:32:18Z
date_updated: 2026-02-17T12:32:18Z
file_id: '21302'
file_name: 2025_AstrophysicalJournal_Kamai.pdf
file_size: 16415089
relation: main_file
success: 1
file_date_updated: 2026-02-17T12:32:18Z
has_accepted_license: '1'
intvolume: ' 994'
language:
- iso: eng
month: '11'
oa: 1
oa_version: Published Version
publication: The Astrophysical Journal
publication_identifier:
eissn:
- 1538-4357
issn:
- 0004-637X
publication_status: published
publisher: IOP Publishing
quality_controlled: '1'
status: public
title: Machine Learning inference of stellar properties using integrated photometric
and spectroscopic data
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
short: CC BY (4.0)
type: journal_article
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 994
year: '2025'
...
---
OA_place: publisher
OA_type: gold
_id: '21327'
abstract:
- lang: eng
text: Proteins exist as a dynamic ensemble of multiple conformations, and these
motions are often crucial for their functions. However, current structure prediction
methods predominantly yield a single conformation, overlooking the conformational
heterogeneity revealed by diverse experimental modalities. Here, we present a
framework for building experiment-grounded protein structure generative models
that infer conformational ensembles consistent with measured experimental data.
The key idea is to treat stateof-the-art protein structure predictors (e.g., AlphaFold3)
as sequence-conditioned structural priors, and cast ensemble modeling as posterior
inference of protein structures given experimental measurements. Through extensive
real-data experiments, we demonstrate the generality of our method to incorporate
a variety of experimental measurements. In particular, our framework uncovers
previously unmodeled conformational heterogeneity from crystallographic densities,
and generates high-accuracy NMR ensembles orders of magnitude faster than the
status quo. Notably, we demonstrate that our ensembles outperform AlphaFold3 (Abramson
et al., 2024) and sometimes better fit experimental data than publicly deposited
structures to the Protein Data Bank (PDB, Burley et al. (2017)). We believe that
this approach will unlock building predictive models that fully embrace experimentally
observed conformational diversity.
acknowledged_ssus:
- _id: ScienComp
acknowledgement: 'This work was supported by the Israeli Science Foundation (ISF)
grant number 1834/24. We acknowledge support from the Austrian Science Fund (FWF,
grant numbers I5812-B and I6223) and the financial support of the Helmsley Fellowships
Program for Sustainability and Health. This research uses resources of the Institute
of Science and Technology Austria’s scientific computing cluster. '
alternative_title:
- PMLR
article_processing_charge: No
arxiv: 1
author:
- first_name: Sai A
full_name: Maddipatla, Sai A
id: e957f5e5-91c9-11f0-a95f-e090f66ecb4d
last_name: Maddipatla
- first_name: Nadav E
full_name: Sellam, Nadav E
id: ef280fe0-91c9-11f0-a95f-8dea3f5bc513
last_name: Sellam
- first_name: Meital I
full_name: Bojan, Meital I
id: 11d88cf5-91ca-11f0-a95f-edf9f08f47b7
last_name: Bojan
- first_name: Sanketh
full_name: Vedula, Sanketh
id: 94f2fe44-70fa-11f0-b76b-92922c09452b
last_name: Vedula
- first_name: Paul
full_name: Schanda, Paul
id: 7B541462-FAF6-11E9-A490-E8DFE5697425
last_name: Schanda
orcid: 0000-0002-9350-7606
- first_name: Ailie
full_name: Marx, Ailie
last_name: Marx
- first_name: Alexander
full_name: Bronstein, Alexander
id: 58f3726e-7cba-11ef-ad8b-e6e8cb3904e6
last_name: Bronstein
orcid: 0000-0001-9699-8730
citation:
ama: 'Maddipatla SA, Sellam NE, Bojan MI, et al. Inverse problems with experiment-guided
AlphaFold. In: Proceedings of the 42nd International Conference on Machine
Learning. Vol 267. ML Research Press; 2025:42366-42393.'
apa: 'Maddipatla, S. A., Sellam, N. E., Bojan, M. I., Vedula, S., Schanda, P., Marx,
A., & Bronstein, A. M. (2025). Inverse problems with experiment-guided AlphaFold.
In Proceedings of the 42nd International Conference on Machine Learning
(Vol. 267, pp. 42366–42393). Vancouver, Canada: ML Research Press.'
chicago: Maddipatla, Sai A, Nadav E Sellam, Meital I Bojan, Sanketh Vedula, Paul
Schanda, Ailie Marx, and Alex M. Bronstein. “Inverse Problems with Experiment-Guided
AlphaFold.” In Proceedings of the 42nd International Conference on Machine
Learning, 267:42366–93. ML Research Press, 2025.
ieee: S. A. Maddipatla et al., “Inverse problems with experiment-guided AlphaFold,”
in Proceedings of the 42nd International Conference on Machine Learning,
Vancouver, Canada, 2025, vol. 267, pp. 42366–42393.
ista: 'Maddipatla SA, Sellam NE, Bojan MI, Vedula S, Schanda P, Marx A, Bronstein
AM. 2025. Inverse problems with experiment-guided AlphaFold. Proceedings of the
42nd International Conference on Machine Learning. ICML: International Conference
on Machine Learning, PMLR, vol. 267, 42366–42393.'
mla: Maddipatla, Sai A., et al. “Inverse Problems with Experiment-Guided AlphaFold.”
Proceedings of the 42nd International Conference on Machine Learning, vol.
267, ML Research Press, 2025, pp. 42366–93.
short: S.A. Maddipatla, N.E. Sellam, M.I. Bojan, S. Vedula, P. Schanda, A. Marx,
A.M. Bronstein, in:, Proceedings of the 42nd International Conference on Machine
Learning, ML Research Press, 2025, pp. 42366–42393.
conference:
end_date: 2025-07-19
location: Vancouver, Canada
name: 'ICML: International Conference on Machine Learning'
start_date: 2025-07-13
corr_author: '1'
date_created: 2026-02-18T12:11:17Z
date_published: 2025-07-30T00:00:00Z
date_updated: 2026-02-19T08:56:43Z
day: '30'
ddc:
- '000'
- '540'
department:
- _id: PaSc
- _id: AlBr
- _id: GradSch
external_id:
arxiv:
- '2502.09372'
file:
- access_level: open_access
checksum: f33230a6d59b7978d4cd72795e4e9059
content_type: application/pdf
creator: dernst
date_created: 2026-02-19T08:56:10Z
date_updated: 2026-02-19T08:56:10Z
file_id: '21338'
file_name: 2025_ICML_Maddipatla.pdf
file_size: 1924177
relation: main_file
success: 1
file_date_updated: 2026-02-19T08:56:10Z
has_accepted_license: '1'
intvolume: ' 267'
language:
- iso: eng
month: '07'
oa: 1
oa_version: Published Version
page: 42366 - 42393
project:
- _id: eb9c82eb-77a9-11ec-83b8-aadd536561cf
grant_number: I05812
name: AlloSpace. The emergence and mechanisms of allostery
- _id: bdb9578d-d553-11ed-ba76-ed5d39fce6f0
grant_number: I06223
name: Structure and mechanism of the mitochondrial MIM insertase
publication: Proceedings of the 42nd International Conference on Machine Learning
publication_identifier:
eissn:
- 2640-3498
publication_status: published
publisher: ML Research Press
quality_controlled: '1'
status: public
title: Inverse problems with experiment-guided AlphaFold
tmp:
image: /images/cc_by.png
legal_code_url: https://creativecommons.org/licenses/by/4.0/legalcode
name: Creative Commons Attribution 4.0 International Public License (CC-BY 4.0)
short: CC BY (4.0)
type: conference
user_id: 2DF688A6-F248-11E8-B48F-1D18A9856A87
volume: 267
year: '2025'
...