Skip to content
DraconiticResearch
Search Sources API
197,768 studies31 sourcesUpdated 20 min ago
Peer-reviewedCC BY-NC-NDNature Communications

Convergence is not correctness: context-dependent performance of enhanced-sampling methods across biological complexity

  • Christopher Kang1
  • Cheng Giuseppe Chen2
  • Chenyu Tang2
  • Sergio Contreras Arredondo2
  • Mengchen Zhou3
  • Haohao Fu3
  • Lan Yang4
  • James C. Gumbart4
  • Ashkan Fakharzadeh5,6,7
  • Mahmoud Moradi8
  • Haochuan Chen7
  • Rui Sun9
3 more authors
  • Jonathan D Harris10
  • Benoı̂t Roux10
  • Christophe J. Chipot11,12,13,14,15
15 affiliations
  1. 1Department of Chemical and Biomolecular Engineering, North Carolina State University, Raleigh, NC, USA
  2. 2Laboratoire de Physique et Chimie Théoriques, Unité Mixte de Recherche no 7019, Université de Lorraine, Vandœuvre-lès-Nancy Cedex, France
  3. 3Research Center for Analytical Sciences, Tianjin Key Laboratory of Biosensing and Molecular Recognition, State Key Laboratory of Medicinal Chemical Biology, College of Chemistry, Nankai University, Tianjin, PR China
  4. 4School of Physics, Georgia Institute of Technology, Atlanta, GA, USA
  5. 5Center for Biophysics and Quantitative Biology, University of Illinois Urbana-Champaign, Urbana, IL, USA
  6. 6Department of Biochemistry, University of Illinois Urbana-Champaign, Urbana, IL, USA
  7. 7Theoretical and Computational Biophysics Group, NIH Resource for Macromolecular Modeling and Visualization, Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, USA
  8. 8Department of Chemistry and Biochemistry, University of Arkansas, Fayetteville, AR, USA
  9. 9Department of Chemistry, The University of Hawai'i at Mānoa, Honolulu, HI, USA
  10. 10Department of Biochemistry and Molecular Biology, The University of Chicago, Chicago, IL, USA
  11. 11Department of Biochemistry and Molecular Biology, The University of Chicago, Chicago, IL, USA. [email protected]
  12. 12Department of Chemistry, The University of Hawai'i at Mānoa, Honolulu, HI, USA. [email protected]
  13. 13Department of Physics, University of Illinois at Urbana-Champaign, Urbana, IL, USA. [email protected]
  14. 14Laboratoire de Physique et Chimie Théoriques, Unité Mixte de Recherche no 7019, Université de Lorraine, Vandœuvre-lès-Nancy Cedex, France. [email protected]
  15. 15Theoretical and Computational Biophysics Group, NIH Resource for Macromolecular Modeling and Visualization, Beckman Institute for Advanced Science and Technology, University of Illinois at Urbana-Champaign, Urbana, IL, USA. [email protected]
Open at sourceDOICite

Abstract

Enhanced-sampling molecular dynamics is indispensable for studying biomolecular processes beyond the reach of conventional simulation, yet its reliability in realistic biological settings remains poorly characterized. Here, we compare four widely used approaches—replica-exchange umbrella sampling (REUS), well-tempered metadynamics (WT-MtD), well-tempered metadynamics–extended adaptive-biasing force (WTM-eABF), and on-the-fly probability enhanced sampling (OPES)—across five experimentally grounded processes: ligand binding, conformational change, kinase activation, membrane permeation, and ion conduction. Three findings emerge. Method performance does not transfer across biological contexts; each approach exhibits characteristic failure modes invisible in low-dimensional model systems but consequential in realistic landscapes. Self-consistency metrics, the field’s standard convergence diagnostic, are insufficient: simulations can satisfy convergence criteria while yielding quantitatively incorrect free-energy profiles. Cross-method agreement against an independent reference is therefore necessary. These results re-frame method selection as a context-dependent decision and argue that cross-method validation should become a community standard. Enhanced-sampling molecular dynamics methods are widely used to probe biomolecular processes, yet their reliability in realistic biological systems remains poorly characterized. This study shows that method performance does not transfer across biological contexts, that standard convergence metrics can be misleading, and that reliable free-energy estimation requires cross-method validation, underscoring the need for context-aware method selection.

Related

  1. N6-Methyladenosine RNA Modifications Influence the Stability of FUS-Dependent Biological Condensate Formation.
    Chamali Thalagaha Mudiyanselage, Milan Pandey, Sudeshi M Abedeera +5
    ACS bio & med chem AuPeer-reviewedCC BY-NC-NDFull textMolecular biology & gene regulationEurope PMC
  2. The African swine fever virus I73R protein is a multifunctional Zα domain that mediates both Z-binding-dependent and -independent biological functions through induction of phase separation
    Sébastien Pirotte, Yingnan Liu, Laurie Conti +12
    Nucleic Acids ResearchPeer-reviewedCC BY-NCFull textMolecular biology & gene regulationCrossrefEurope PMCPMC
  3. Epitope Generation Gateway: A biological context-dependent approach to cancer vaccine development
    Luca Mannino, Michele Fratello, Jacopo Chiaro +8
    GigaSciencePeer-reviewedCC BY 4.0Full textMolecular biology & gene regulationCrossref
  4. Plastic mulching consistently improves brinjal performance, but drip irrigation benefits are soil dependent: A two-location field study
    Md. Nayan, F. M. Jamil Uddin, Rounak Islam +5
    PLOS OnePeer-reviewedCC BY 4.0Full textGeneral biochemistryPlant biologyPLOSPMC
  5. Dimensional convergence at the nanoscale: a novel synthesis of experimental evidence justifying direct measurement of quantum coherence in microtubules
    Justin Echternach
    Frontiers in molecular biosciencesPeer-reviewedCC BYFull textMolecular biology & gene regulationEurope PMC

Cite

BibTeXRIS
BibTeX
@article{kang2026convergence,
  title = {{Convergence is not correctness: context-dependent performance of enhanced-sampling methods across biological complexity}},
  author = {Christopher Kang and Cheng Giuseppe Chen and Chenyu Tang and Sergio Contreras Arredondo and Mengchen Zhou and Haohao Fu and Lan Yang and James C. Gumbart and Ashkan Fakharzadeh and Mahmoud Moradi and Haochuan Chen and Rui Sun and Jonathan D Harris and Benoı̂t Roux and Christophe J. Chipot},
  journal = {Nature Communications},
  year = {2026},
  month = jun,
  doi = {10.1038/s41467-026-74165-y},
  url = {https://doi.org/10.1038/s41467-026-74165-y},
  note = {CC BY-NC-ND},
}
RIS
TY  - JOUR
TI  - Convergence is not correctness: context-dependent performance of enhanced-sampling methods across biological complexity
AU  - Christopher Kang
AU  - Cheng Giuseppe Chen
AU  - Chenyu Tang
AU  - Sergio Contreras Arredondo
AU  - Mengchen Zhou
AU  - Haohao Fu
AU  - Lan Yang
AU  - James C. Gumbart
AU  - Ashkan Fakharzadeh
AU  - Mahmoud Moradi
AU  - Haochuan Chen
AU  - Rui Sun
AU  - Jonathan D Harris
AU  - Benoı̂t Roux
AU  - Christophe J. Chipot
PY  - 2026
DA  - 2026/06/12
JO  - Nature Communications
DO  - 10.1038/s41467-026-74165-y
UR  - https://doi.org/10.1038/s41467-026-74165-y
AB  - Enhanced-sampling molecular dynamics is indispensable for studying biomolecular processes beyond the reach of conventional simulation, yet its reliability in realistic biological settings remains poorly characterized. Here, we compare four widely used approaches—replica-exchange umbrella sampling (REUS), well-tempered metadynamics (WT-MtD), well-tempered metadynamics–extended adaptive-biasing force (WTM-eABF), and on-the-fly probability enhanced sampling (OPES)—across five experimentally grounded processes: ligand binding, conformational change, kinase activation, membrane permeation, and ion conduction. Three findings emerge. Method performance does not transfer across biological contexts; each approach exhibits characteristic failure modes invisible in low-dimensional model systems but consequential in realistic landscapes. Self-consistency metrics, the field’s standard convergence diagnostic, are insufficient: simulations can satisfy convergence criteria while yielding quantitatively incorrect free-energy profiles. Cross-method agreement against an independent reference is therefore necessary. These results re-frame method selection as a context-dependent decision and argue that cross-method validation should become a community standard. Enhanced-sampling molecular dynamics methods are widely used to probe biomolecular processes, yet their reliability in realistic biological systems remains poorly characterized. This study shows that method performance does not transfer across biological contexts, that standard convergence metrics can be misleading, and that reliable free-energy estimation requires cross-method validation, underscoring the need for context-aware method selection.
SN  - 2041-1723
ER  -