Journal Article Epublish
PubMed: 40177350 DOI: 10.3389/fchem.2025.1395374
Development of the TSR-based computational method to investigate spike and monoclonal antibody interactions.
Front Chem 13:1395374
Authors: Tarikul I Milon, Titli Sarkar, Yixin Chen, Jordan M Grider, Feng Chen, Jun-Yuan Ji, Seetharama D Jois, Konstantin G Kousoulas, Vijay Raghavan, Wu Xu
Affiliations
- Department of Chemistry, University of Louisiana at Lafayette, Lafayette, LA, United States.
- The Center for Advanced Computer Studies, University of Louisiana at Lafayette, Lafayette, LA, United States.
- Department of Computer and Information Science, The University of Mississippi, University, MS, United States.
- High Performance Computing, 329 Frey Computing Services Center, Louisiana State University, Baton Rouge, LA, United States.
- Department of Biochemistry and Molecular Biology, Tulane University School of Medicine, Louisiana Cancer Research Center, New Orleans, LA, United States.
- Department of Pathobiological Sciences, LSU School of Veterinary Medicine, Louisiana State University, Baton Rouge, LA, United States.
Abstract
Introduction
Monoclonal antibody (mAb) drug treatments have proven effective in reducing COVID-19-related hospitalizations or fatalities, particularly among high-risk patients. Numerous experimental studies have explored the structures of spike proteins and their complexes with ACE2 or mAbs. These 3D structures provide crucial insights into the interactions between spike proteins and ACE2 or mAb, forming a basis for the development of diagnostic tools and therapeutics. However, the field of computational biology has faced substantial challenges due to the lack of methods for precise protein structural comparisons and accurate prediction of molecular interactions. In our previous studies, we introduced the Triangular Spatial Relationship (TSR)-based algorithm, which represents a protein's 3D structure using a vector of integers (keys). These earlier studies, however, were limited to individual proteins.
Purpose
This study introduces new extensions of the TSR-based algorithm, enhancing its ability to study interactions between two molecules. We apply these extensions to gain a mechanistic understanding of spike - mAb interactions.
Method
We expanded the basic TSR method in three novel ways: (1) TSR keys encompassing all atoms, (2) cross keys for interactions between two molecules, and (3) intra-residual keys for amino acids. This TSR-based representation of 3D structures offers a unique advantage by simplifying the search for similar substructures within structural datasets.
Results
The study's key findings include: (i) The method effectively quantified and interpreted conformational changes and steric effects using the newly introduced TSR keys. (ii) Six clusters for CDRH3 and three clusters for CDRL3 were identified using all-atom keys. (iii) We constructed the TSR-STRSUM (TSR-STRucture SUbstitution Matrix), a matrix that represents pairwise similarities between amino acid structures, providing valuable applications in protein sequence and structure comparison. (iv) Intra-residual keys revealed two distinct Tyr clusters characterized by specific triangle geometries.
Conclusion
This study presents an advanced computational approach that not only quantifies and interprets conformational changes in protein backbones, entire structures, or individual amino acids, but also facilitates the search for substructures induced by molecular binding across protein datasets. In some instances, a direct correlation between structures and functions was successfully established.
MeSH
Amino Acids, Algorithms, Machine Learning, Antibodies, Monoclonal, Therapeutics