Transcript Statistical analysis of DNA microarray data

Network Biology
BMI 730
Kun Huang
Department of Biomedical Informatics
Ohio State University
Systems Sciences
Understanding!
Theory
Analysis
Modeling
• Synthesis/prediction
• Simulation
• Hypothesis generation
Prediction!
Systems Biology
Biology
Informatics
Domain knowledge
Data management
• Hypothesis testing
Experimental work
• Genetic manipulation
• Quantitative measurement
• Validation
• Database
Computational infrastructure
• Modeling tools
• High performance computing
Visualization
Review of Network Topology – Scale
Free and Modularity
Elements of Dynamical Modeling
Network Motif Analysis
Integration of Multiple Networks –
Several Examples
Course Projects
A Tale of Two Groups
A.-L. Barabasi at University of Notre Dame
Ten Most Cited Publications:
Albert-László Barabási and Réka Albert, Emergence of scaling in random networks , Science 286, 509512 (1999). [ PDF ] [ cond-mat/9910332 ]
Réka Albert and Albert-László Barabási, Statistical mechanics of complex networks
Review of Modern Physics 74, 47-97 (2002). [ PDF ] [cond-mat/0106096 ]
H. Jeong, B. Tombor, R. Albert, Z.N. Oltvai, and A.-L. Barabási, The large-scale organization of
metabolic networks, Nature 407, 651-654 (2000). [ PDF ] [ cond-mat/0010278 ]
R. Albert, H. Jeong, and A.-L. Barabási, Error and attack tolerance in complex networks
Nature 406 , 378 (2000). [ PDF ] [ cond-mat/0008064 ]
R. Albert, H. Jeong, and A.-L. Barabási, Diameter of the World Wide Web
Nature 401, 130-131 (1999). [ PDF ] [ cond-mat/9907038 ]
H. Jeong, S. Mason, A.-L. Barabási and Zoltan N. Oltvai, Lethality and centrality in protein networks
Nature 411, 41-42 (2001). [ PDF ] [ Supplementary Materials 1, 2 ]
E. Ravasz, A. L. Somera, D. A. Mongru, Z. N. Oltvai, and A.-L. Barabási, Hierarchical organization of
modularity in metabolic networks, Science 297, 1551-1555 (2002). [ PDF ] [ cond-mat/0209244 ] [
Supplementary Material ]
A.-L. Barabási, R. Albert, and H. Jeong, Mean-field theory for scale-free random networks
Physica A 272, 173-187 (1999). [ PDF ] [ cond-mat/9907068 ]
Réka Albert and Albert-László Barabási, Topology of evolving networks: Local events and universality
Physical Review Letters 85, 5234 (2000). [ PDF ] [ cond-mat/0005085 ]
Albert-László Barabási and Zoltán N. Oltvai, Network Biology: Understanding the cells's functional
organization, Nature Reviews Genetics 5, 101-113 (2004). [ PDF ]
Power Law
Small World
Rich Get Richer
(preferential
attachment)
Self-similarity
HUBS!
Modularity
Scale-free and Modularity/Hierarchy are thought to be
exclusive.
(a)
Scale-free
(b)
Modular
Subgraphs
•
•
Subgraph: a connected graph
consisting of a subset of the nodes and
links of a network
Subgraph properties:
n: number of nodes
m: number of links
(n=3,m=3)
(n=3,m=2)
(n=4,m=4)
(n=4,m=5)
.
R Milo et al., Science 298, 824-827 (2002).
Review of Network Topology – Scale
Free and Modularity
Elements of Dynamical Modeling
Network Motif Analysis
Integration of Multiple Networks –
Several Examples
Course Projects
Genetic Network – Transcription Network
• Regulation of protein expression is mediated by
transcription factors
Promoter
DNA
Gene Y
Protein Y
Translation
RNA polymerase
mRNA
Transcription
DNA
Gene Y
Genetic Network – Transcription Network
• TF factor X regulates protein (gene) Y
Y
Protein Y
DNA
Y
Y
Y
Y
mRNA
SX
X
Y
X*
X*
Gene Y
Activation / positive control, X is called activator.
XY
Genetic Network – Transcription Network
• TF factor X regulates protein (gene) Y
X
X*
No transcription
DNA
X*
Gene Y
Y
mRNA
Y
Y
Y
X
DNA
Gene Y
Repression / negative control, X is called repressor.
X
Y
Genetic Network – Transcription Network
• How to model the input-output relationship?
Concentration of
active TF X*
Rate of production
of protein Y
Concentration of
protein Y
F(X*) is usually monotonic, S-shaped function.
Genetic Network – Transcription Network
• Hill function
• Derived from the equilibrium binding of the TF to its target
site.
Activator
K – activation coefficient
 – maximal expression level
n – Hill coefficient (1<n<4 for most cases)
F(X*) approximates step function (logic) for large n

n=4
n=2
n=1
/2
0
1
X*/K
X*>>K, F(X*) = 
X* = K, F(X*) = /2
2
Genetic Network – Transcription Network
Repressor
F(X*) approximates step function (logic) for large n

n=2
n=4
n=1
/2
0
1
X*/K
2
Genetic Network – Transcription Network
• TF factor X regulates protein (gene) Y
• Timescale for E. Coli
1. Binding of signaling molecule to TF and
changing its activity
~1msec
2. Binding of active TF to DNA
~1sec
3. Transcription + translation of gene
~5min
4. 50% change of target protein concentration
~1h
Genetic Network – Transcription Network
• Logic function approximation
• Hill function is for detailed modeling. Logic
function is for simplicity and mathematical clarity.
q
0
t
Activator
Repressor
K – threshold
 – maximal expression level
Genetic Network – Transcription Network
• Logic function approximation
• Multiple input
X* AND Y*
X* OR Y*
SUM
Genetic Network – Transcription Network
•
•
•
•
•
The dynamics
Change over time
Degradation
Dilution (cell growth and volume increase)
Response time (characteristics)
Dynamical equation
Equilibrium (steady state)
Genetic Network – Transcription Network
• The dynamics
• Response time (characteristics)
• Sudden removal of production
0.5
1
Genetic Network – Transcription Network
• The dynamics
• Response time (characteristics)
• Sudden initiation of production
0.5
1
Motif Statistics and Dynamics
• Autoregulation
• Self-edge in the transcription network
Motif Statistics and Dynamics
• Autoregulation
Negative autoregulation
X
A
mRNA
DNA
Gene Y
Motif Statistics and Dynamics
• Autoregulation
X
A
mRNA
DNA
X(t)/K
Gene Y
1
0
Time (at)
1
X(t)/K
Motif Statistics and Dynamics
• Autoregulation
1
0
Time (at)
Short response time
1
Motif Statistics and Dynamics
• Autoregulation
If  fluctuates, Xss is stable for negative
autoregulation but not for simple regulation.
Robustness / stabilization
Review of Network Topology – Scale
Free and Modularity
Elements of Dynamical Modeling
Network Motif Analysis
Integration of Multiple Networks –
Several Examples
Course Projects
Motif Topology
Each edge has 4 choices (why?).
Three edges 4X4X4 = 64
choices. There are symmetry
redundancy. Despite the choices
of activation and repression,
there are 13 types.
Coherent Feed Forward Loop (FFL)
X
X
X
X
Y
Y
Y
Y
Z
Z
Z
Z
Incoherent Feed Forward Loop
X
X
X
X
Y
Y
Y
Y
Z
Z
Z
Z
Coherent Feed Forward Loop (FFL)
Sx
X
Y
Z
Sx
X
Y
Ton
AND
Z
Sign sensitive delay for ON signal
Coherent Feed Forward Loop (FFL)
Sx
X
Y
Z
Sx
X
Y
AND
Z
Sign sensitive delay for ON signal
Coherent Feed Forward Loop (FFL)
The Coherent Feedforward Loop Serves as a Sign-sensitive Delay Element in Transcription Networks
Mangan, S.; Zaslaver, A.; Alon, U. J. Mol. Biol., 334:197-204, 2003.
Coherent Feed Forward Loop (FFL)
Timing instrument
Coherent Feed Forward Loop (FFL)
X
Y
Z
Sx
Sy
X
Y
AND
Z
Nature Genetics 31, 64 - 68 (2002)
Network motifs in the transcriptional regulation network of Escherichia coli
Shai S. Shen-Orr, Ron Milo, Shmoolik Mangan & Uri Alon
Noise (low-pass) filter
Coherent Feed Forward Loop (FFL)
Sx
X
Y
Z
Sx
X
Y
OR
Z
Sign sensitive delay for OFF signal
Coherent Feed Forward Loop (FFL)
A coherent feed-forward loop with a SUM input
function prolongs flagella expression in
Escherichia coli
Shiraz Kalir, Shmoolik Mangan and Uri Alon, Mol. Sys. Biol., Mar.2005.
Coherent Feed Forward Loop (FFL)
A coherent feed-forward loop with a SUM input
function prolongs flagella expression in
Escherichia coli
Shiraz Kalir, Shmoolik Mangan and Uri Alon, Mol. Sys. Biol., Mar.2005.
Incoherent Feed Forward Loop (FFL)
X
Sx
Y
Z
Sx
X
Y
AND
Z
Fast response time to steady state
Table 3. Summary of functions of the FFLs
*
Steady-state logic is sensitive to both
Sx and Sy
Coherent and incoherent*
Types 1, 2
AND
Types 3, 4
OR
Sign-sensitive delay upon Sx steps
Coherent
Types 1, 2,
3, 4
Sy-gated pulse generator upon Sx
steps
Incoherent with no basal
Y level
Types 3, 4
AND
Types 1,2
OR
Sign-sensitive acceleration upon Sx
steps
Incoherent with basal Y
level
Types
1,2,3,4
In incoherent FFL with basal level, Sy modulates Z between two nonzero levels.
Mangan, S. and Alon, U. (2003) Proc. Natl. Acad. Sci. USA 100, 11980-11985
Review of Network Topology – Scale
Free and Modularity
Elements of Dynamical Modeling
Network Motif Analysis
Integration of Multiple Networks –
Several Examples
Course Projects
Integration of Multi-Modal Data
Barabasi A-L,
Network medicine
– from obesity to
“Diseasome”,
NEJM, 357(4): 404407, 2007.
Tissue-Tissue Network
Dobrin et al. Genome Biology 2009
10:R55 doi:10.1186/gb-2009-10-5-r55
Tissue-Tissue Network
Dobrin et al. Genome Biology 2009
10:R55 doi:10.1186/gb-2009-10-5-r55
Genotype-Phenotype Network
Scoring scheme of CIPHER. First, the human phenotype network, protein network, and gene–
phenotype network are assembled into an integrated network. Then, to score a particular phenotype–
gene pair (p, g), the phenotype similarity profile for p is extracted and the gene closeness profile for g
is computed from the integrated network. Finally, the linear correlation of the two profiles is calculated
and assigned as the concordance score between the phenotype p and the gene g.
Wu et al. Molecular Systems Biology, 2009 4:189, Network-based global inference
of human disease genes
Genotype-Phenotype Network
Known
disease
gene
BRCA1
AR
ATM
CHEK2
BRCA2
STK11
RAD51
PTEN
BARD1
TP53
RB1CC1
NCOA3
PIK3CA
PPM1D
CASP8
TGF1
Rank in 8919 candidates
CIPHER-SP
%
CIPHER-DN
%
1
3
19
66
139
150
174
188
196
287
798
973
1644
1946
4978
7116
0.01
0.03
0.21
0.74
1.56
1.69
2.00
2.10
2.20
3.22
8.95
10.91
18.43
21.82
55.81
79.78
2
3
4
19
49
21
36
24
41
45
6360
343
367
7318
2397
3502
0.02
0.03
0.04
0.21
0.54
0.23
0.40
0.26
0.45
0.50
71.30
3.84
4.11
82.04
26.87
39.26
Wu et al. Molecular Systems Biology, 2009 4:189, Network-based global inference
of human disease genes
Kelley and Ideker, Nature Biotechnology, 2005 23:561-566, Systematic
interpretation of genetic interactions using protein networks
Review of Network Topology – Scale
Free and Modularity
Elements of Dynamical Modeling
Network Motif Analysis
Integration of Multiple Networks –
Several Examples
Course Projects