introduction to bioinformatics - ucl computer science · introduction to bioinformatics biological...
TRANSCRIPT
Introduction to Bioinformatics
Biological Networks
Department of ComputingImperial College London
Spring 2010
Nataša Prž[email protected]
• Large Networks model many real-world phenomena– technological: www, internet, electric circuits,...
– social: friendship, collaboration, disease spread,...
– biological: • protein structure,
• transcriptional regulation,
• metabolic,
• protein-protein interaction (PPI),
• …
1. Motivation
Nataša Prž[email protected]
2
1. Motivation
• Large Networks model many real-world phenomena– technological: www, internet, electric circuits,...
– social: friendship, collaboration, disease spread,...
– biological: • protein structure,
• transcriptional regulation,
• metabolic,
• protein-protein interaction (PPI),
• …
Nataša Prž[email protected]
3
1. Motivation
• Large Networks model many real-world phenomena– technological: www, internet, electric circuits,...
– social: friendship, collaboration, disease spread,...
– biological: • protein structure,
• transcriptional regulation,
• metabolic,
• protein-protein interaction (PPI),
• …
Nataša Prž[email protected]
4
1. Motivation
• Large-scale networks in bioinformatics:– Technological advances in experimental biology
data
– Important computational problems
– Algorithmic and modeling advances contribute:• biological understanding (function, disease, pathogens,…)
• therapeutics
Booming research area
Nataša Prž[email protected]
5
• Why model biological networks?– Concise summary, unexpected properties– Understand laws predictions/reproduction
• E.g. – Johannes Kepler (1571-1630)
» Observed planetary motion– Sir Isaac Newton (1643-1727)
» Universal gravitation, laws of motion Explained planetary motion
1. Motivation
Nataša Prž[email protected]
6
Problems:
1. Noise revise models as data sets evolve2. “Hardness” of graph theoretic problems
E.g. NP-completeness of subgraph isomorphism Cannot exactly compare/align networks heuristics (approximate solutions)
Exact comparison inappropriate in biology due to biological variation
1. Motivation
Nataša Prž[email protected]
7
1. MotivationProperties of Large Networks (heuristic comparisons)•Global
•Degree distribution•Diameter•Clustering coefficient/spectrum
•Local:•network “motifs” and subgraphs
(U. Alon’s group, ’02-’04, Przulj 2004)
Nataša Prž[email protected]
8
262626262626
Assumed Knowledge of Graph Theory and Algorithms:
Nataša Prž[email protected]
A program for testing isomorphism and automorphism of graphs:Brendan McKay’s “nauty”: http://cs.anu.edu.au/~bdm/nauty/
27272727
Assumed Knowledge of Graph Theory and Algorithms:
Nataša Prž[email protected]
Example:
abc
Node a can be mapped to c by an automorphism,and b can only be mapped to itself.
Thus: Orb(a)={a,c}, Orb(b)={b}.
28
Assumed Knowledge of Graph Theory and Algorithms:
Nataša Prž[email protected]
- Pseudocode (Chpt 1.1)- Growth Rate of Running Time
by Goodrich and Tamassia