Network Analysis : Part 2 , Agent-Based Modeling on Protein Networks for Drug Discovery

My previous post i used Proximity analysis reveal how closely drug targets relate to disease-associated genes, suggesting potential therapeutic applications and drug repurposing opportunities. We used null models, specifically degree-matching and degree log-binning, to create randomized networks to understand Largest connected components. We also used proximity analysis to compute average shortest path length between disease-associated genes and drug targets .

This post is an extension of the previous post and goes in depth with touch to network analysis and use of Agent-Based Models (ABMs). ABMs serves as a powerful toolkit for examining natural phenomena, providing a means to encode both established knowledge and new theories within a rigorous quantitative framework. By aligning these models with experimental observations and using them to predict how a system might behave under previously untested scenarios, researchers can gain fresh insights and develop innovative hypotheses. Questions like “Which molecular properties are essential?”, “What dose might be suitable in humans?”, “Where do our uncertainties lie?”, and “How should the next experiment be structured?” can largely be addressed by building and interrogating carefully designed mathematical models.

Unlike purely analytical or equation-based approaches, ABMs let us track individual “agents” (which can be proteins or cells) in a virtual environment where they follow a set of…