How to map a protein structure into a network graph [...]
summary
Method converts atomic coordinates into characteristic vectors for secondary elements, builds a complete weighted graph (nodes = vectors, edges = distance/angle metrics), and uses Python-igraph plus graph-product and spanning-tree filters to find a maximum common core for 3D alignment. It derives eigenvector centrality to map interface vs buried residues, scales to batch PDB processing with sampling, and reports it works for folded proteins but does not handle disordered proteins and needs downsampling for very large complexes.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 2:51What is a protein structure
- 8:53Protein structures as graphs
- 12:14Building graphs from vectors
- 17:48Comparing structures with graphs
- 25:18Analyzing interfaces with centrality
- 28:30Conclusion and future apps
- 29:59Q&A
related talks
16mGraph Network in High Energy Physics
24mNetworkX - your Unexpected Assistant for Clustering Analysis
27mWhat Is the Most Important Part of Your Graph? Understanding Centrality and GDS
01h 26mBuilding an Interactive Network Graph to Understand Communities
01h 38mNetwork Science for Fun and Profit
01h 33mBuilding an Interactive Network Graph to Understand Communities
37mGraphs, Networks and Python: The Power of Interconnection
26mGetting an Edge with Network Analysis with Python