Skip to content

Chapter 16: Systems Biology: Integrating the 'Omics'

šŸ“„ Download Lecture Slides (PPTX)

16.1 Reductionism vs. Holism

Traditional biology is reductionist: it breaks things down to study them (e.g., studying one gene at a time).

Systems Biology is holistic: it studies the interactions between the parts. It views the cell not as a bag of individual parts, but as a complex, interconnected network.

"The whole is greater than the sum of its parts."

16.2 Biological Networks

Illustration of Biological Networks

15.4 Network Analysis and Pathway Enrichment

Systems biology connects molecular entities into networks and pathways. Practical components:

  • Network construction: Build co-expression or proteinā€“protein interaction networks (WGCNA, STRING interactions).
  • Module detection: Identify modules/clusters and test for enrichment of pathways or gene ontology terms.
  • Pathway enrichment: Tools include g:Profiler, ReactomePA, clusterProfiler for pathway over-representation and GSEA approaches.

Best practices:

  • Use multiple evidence sources (expression, interaction databases) to build robust networks.
  • Correct for multiple testing and report effect sizes, not only p-values.
  • Visualize modules with clear legends, colors, and interactive viewers where possible.

We represent these systems using Graphs (Networks). * Nodes: The biological entities (Genes, Proteins, Metabolites). * Edges: The relationships (Interacts with, Regulates, Converts to).

Types of Networks

  1. Protein-Protein Interaction (PPI) Networks: Who talks to whom? (Physical binding).
  2. Gene Regulatory Networks (GRN): Who is the boss? (Transcription factors controlling gene expression).
  3. Metabolic Networks: The factory floor. (Enzymes converting Substrate A to Product B).

16.3 Network Topology: Hubs and Bottlenecks

Biological networks are not random. They are Scale-Free Networks. * Most nodes have very few connections. * A few nodes (Hubs) have a massive number of connections.

TP53 (the guardian of the genome) is a classic hub protein. If you knock out a random gene, the cell might survive. If you knock out a hub like TP53, the system collapses (often leading to cancer).

16.4 Bioinformatics in Action: Network Analysis

We use the Python library networkx to analyze these graphs.

import networkx as nx

# Create an empty graph
G = nx.Graph()

# Add interactions (Edges)
# Imagine Protein A interacts with B, C, and D (A is a Hub)
interactions = [
    ("Protein_A", "Protein_B"),
    ("Protein_A", "Protein_C"),
    ("Protein_A", "Protein_D"),
    ("Protein_B", "Protein_C"),
    ("Protein_E", "Protein_F")
]

G.add_edges_from(interactions)

# Calculate "Degree Centrality" (How connected is each node?)
centrality = nx.degree_centrality(G)

# Find the most connected node
most_connected = max(centrality, key=centrality.get)

print("Network Nodes:", G.nodes())
print(f"The Hub is: {most_connected} with score {centrality[most_connected]:.2f}")

Output: 

Network Nodes: ['Protein_A', 'Protein_B', 'Protein_C', 'Protein_D', 'Protein_E', 'Protein_F']
The Hub is: Protein_A with score 0.60

15.5 Integration

The ultimate goal of systems biology is to integrate data from Genomics, Transcriptomics, Proteomics, and Metabolomics into a single model. This allows us to simulate entire cells (e.g., Flux Balance Analysis) to predict how a bacteria will grow or how a drug will affect a human cell.

Summary

Systems Biology puts the pieces of the puzzle back together. By analyzing networks and identifying hubs, we understand the organization of life.