EUVIMEDScopri EUVIMED
ÜbersichtsarbeitOpen access · CC BY 4.0

Molecular Adaptations of Thermophilic Proteins—From Mechanistic Understanding to Machine Learning Approaches

Sermarajan Arunachalam, Balamurali Mm iD, Ramachandran Gnanasekaran iD

ChemistryOpen · 2026-07 · 15 (7)

DOI: 10.1002/open.70234 ↗

At a glance

Übersichtsarbeit in ChemistryOpen, published online on 2026-07-13. Written by Sermarajan Arunachalam and 2 co-authors. Institution: Department of Chemistry, School of Advanced Sciences, Vellore Institute of Technology, Chennai, India. Major MEDLINE topics: Machine Learning, Proteins. 121 references. The full text is freely available (open access, CC BY 4.0).

AbstractSubjectsDati bibliograficiCiteRelated studies

Abstract (original language)

Thermophilic proteins are known for their stability and activity at elevated temperatures and hence serve as useful models for engineering enzymes with enhanced stability. The stability of these proteins is governed by both thermodynamic and kinetic factors that affect the folding–unfolding equilibrium and the rate of irreversible denaturation. Structural studies have shown that enhanced thermostability often results from multiple cooperative interactions, including improved hydrophobic packing, strengthened electrostatic interactions, and optimized hydrogen‐bonding networks. Comparative studies between thermophilic and mesophilic proteins have helped to identify general mechanisms that support thermal adaptations. Mutational and protein‐engineering approaches have further demonstrated the influence of specific amino acid residues or motifs on the stability through altered or modified local structural interactions. In recent years, computational approaches and machine learning methods have emerged as inevitable tools to predict protein stability. Although these approaches have improved the predictions, there still exist several limitations like limited training datasets, bias toward certain protein families, difficulties in model interpretation, etc. Therefore, combining experimental thermodynamic data, structural studies, and modern AI‐based computational tools, it is possible to develop more reliable strategies to design thermostable proteins, in particular, for various industrial and biotechnological applications.

Subjects

MeSH terms (MEDLINE)

Machine LearningProteinsThermodynamicsProtein StabilityProtein EngineeringProtein FoldingHydrophobic and Hydrophilic InteractionsHydrogen Bonding

Author keywords

artificial intelligencedeep learningmachine learningmesophilesmutational studiesprotein engineeringprotein foldingthermophiles

Substances

Proteins

Authors and institutions

  • Department of Chemistry, School of Advanced Sciences, Vellore Institute of Technology, Chennai, India.
  • Centre for Healthcare Advancement, Innovation and Research, Vellore Institute of Technology, Chennai, India.

ORCID profiles: Balamurali Mm, Ramachandran Gnanasekaran

Cite

APA

Arunachalam, S., Mm, B., & Gnanasekaran, R. (2026). Molecular Adaptations of Thermophilic Proteins—From Mechanistic Understanding to Machine Learning Approaches. ChemistryOpen, 15(7). https://doi.org/10.1002/open.70234

Vancouver

Arunachalam S, Mm B, Gnanasekaran R. Molecular Adaptations of Thermophilic Proteins—From Mechanistic Understanding to Machine Learning Approaches. ChemistryOpen. 2026;15(7). doi:10.1002/open.70234

BibTeX
@article{Arunachalam2026,
  author = {Arunachalam, Sermarajan and Mm, Balamurali and Gnanasekaran, Ramachandran},
  title = {Molecular Adaptations of Thermophilic Proteins—From Mechanistic Understanding to Machine Learning Approaches},
  journal = {ChemistryOpen},
  year = {2026},
  volume = {15},
  number = {7},
  doi = {10.1002/open.70234}
}
Start for free →