Using Persistent Homology for Topological Analysis of Protein Interaction Network of Candida Antarctica Lipase B Molecular Dynamic Simulation Model

In this work, we aim to examine the activity of one of the most efficient and commonly used lipases, Candida Antarctica Lipase B (CalB), from the perspective of multiple computational techniques. To this end, we first conduct a series of Molecular Dynam- ics Simulations on CalB in different conditions to analyze the conformational changes of the protein and probe its unusual high-temperature activity. Next, we build the protein interaction network of amino acids for CalB to study pairwise interactions between amino acids (nodes) and probe the protein in terms of statistical features of links’ distribution. Finally, we employ an algebraic topology-based method to study the protein interaction network from a broader perspective. The ”Persistent Homol- ogy (PH) method” is then presented as a way to exceed pairwise interactions and examine protein networks in terms of patterns of interaction between the nodes. Per- sistent Homology studies the evolution of the protein interaction network’s topologi- cal features (homology groups) in different states. Employing topological analysis, we compare the active form of CalB at high temperatures to its inactive states to account for possible topological contributions to the protein functionality. By discovering a prominent 1-dimensional hole in the active form of the protein, we highlight the role of higher-order interaction patterns in the network. Moreover, using the evolution of topological features, we study topological changes in protein networks and show the decline in the total number of 1-dimensional features as the protein loses activity and compactness over time. Accordingly, we propose that the protein’s general conforma- tional changes and three-dimensional structure are not the only facets contributing to its active state. Instead, we suggest examining the topology of the protein inter- action network, referred to as different dimensional holes of the networks, as a higher dimensional analysis should be used to account for protein functionality. Hence, in this work, we desire to present that one needs to consider topological features acting as patterns of interaction between the components to study, examine or predict the folding of polypeptide chains into active structures.

Name thesis:degree_name

Ph.D. Physics

Level thesis:degree_level

Doctoral

Discipline thesis:degree_discipline

Faculty of Mathematics and Science

Department dc:contributor.department

Department of Physics

Grantor

Brock University

Year dc:date.issued

2023

Author dc:creator

• Tajik, Samin

• Persistent Homology, Topological Data Analysis, Protein NEtwork

Statement dc:rights

• CC0 1.0 Universal

Licence dc:rights.uri

• http://creativecommons.org/publicdomain/zero/1.0/

Language dc:language.iso

eng

Handle dc:identifier.uri

http://hdl.handle.net/10464/17289

OAI identifier oai:identifier

oai:brocku.scholaris.ca:10464/17289