CG920 Genomics
Lesson 11
Systems Biology
Jan Hejátko
Functional Genomics and Proteomics of Plants,
CEITEC - Central European Institute of Technology
And
National Centre for Bimolecular Research,
Faculty of Science,
Masaryk University, Brno
hejatko@sci.muni.cz, www.ceitec.eu
2
 Literature sources for Chapter 12:
 Wilt, F.H., and Hake, S. (2004). Principles of Developmental Biology. (New York ; London: W. W.
Norton)
 Eden, E., Navon, R., Steinfeld, I., Lipson, D., and Yakhini, Z. (2009). GOrilla: a tool for discovery
and visualization of enriched GO terms in ranked gene lists. BMC Bioinformatics 10, 48.
 The Arabidopsis Genome Initiative. (2000). Analysis of the genome sequence of the flowering
plant Arabidopsis thaliana. Nature 408, 796-815.
 Benitez, M. and Hejatko, J. Dynamics of cell-fate determination and patterning in the vascular
bundles of Arabidopsis thaliana (submitted)
 de Luis Balaguer MA, Fisher AP, Clark NM, Fernandez-Espinosa MG, Moller BK, Weijers D,
Lohmann JU, Williams C, Lorenzo O, Sozzani R. 2017. Predicting gene regulatory networks by
combining spatial and temporal gene expression data in Arabidopsis root stem cells. Proc Natl
Acad Sci U S A 114(36): E7632-E7640.
Literature
3
 Definition of Systems Biology
 Tools
 Gene Ontology Analysis
 Bayesian Networks
 Molecular/Gene Regulatory Networks Modeling
 Inferring Gene Regulatory Networks from Large Omics
Datasets
Outline
4
Systems biology is the computational and mathematical
analysis and modeling of complex biological systems. It
is a biology-based interdisciplinary field of study that
focuses on complex interactions within biological
systems, using a holistic approach (holism instead of the
more traditional reductionism) to biological research
(Wikipedia).
Definition
5
Systems biology is the study of biological systems
whose behaviour cannot be reduced to the linear sum of
their parts’ functions. Systems biology does not
necessarily involve large numbers of components or vast
datasets, as in genomics or connectomics, but often
requires quantitative modelling methods borrowed from
physics (Nature).
Definition
6
Nice explanatory video by Dr. Nathan Price,
associate director of the Institute for Systems Biology at
https://www.youtube.com/watch?v=OrXRl_8UFHU.
Definition
7
 Definition of Systems Biology
 Tools
 Gene Ontology analysis
Outline
8
Results of –omics Studies vs
Biologically Relevant Conclusions
□ Results of –omics studies represent huge amount of data,
e.g. genes with differential expression. But how to get any
biologically relevant conclusions out of it?
gene locus sample_1 sample_2 status value_1 value_2 log2(fold_change) test_stat p_value q_value significant
AT1G07795 1:2414285-2414967 WT MT OK 0 1,1804 1.79769e+308
1.79769e+3
08 6.88885e-05
0,00039180
1 yes
HRS1 1:4556891-4558708 WT MT OK 0 0,696583 1.79769e+308
1.79769e+3
08 6.61994e-06
4.67708e-
05 yes
ATMLO14 1:9227472-9232296 WT MT OK 0 0,514609 1.79769e+308
1.79769e+3
08 9.74219e-05
0,00053505
5 yes
NRT1.6 1:9400663-9403789 WT MT OK 0 0,877865 1.79769e+308
1.79769e+3
08 3.2692e-08
3.50131e-
07 yes
AT1G27570 1:9575425-9582376 WT MT OK 0 2,0829 1.79769e+308
1.79769e+3
08 9.76039e-06 6.647e-05 yes
AT1G60095 1:22159735-22162419 WT MT OK 0 0,688588 1.79769e+308
1.79769e+3
08 9.95901e-08
9.84992e-
07 yes
AT1G03020 1:698206-698515 WT MT OK 0 1,78859 1.79769e+308
1.79769e+3
08 0,00913915 0,0277958 yes
AT1G13609 1:4662720-4663471 WT MT OK 0 3,55814 1.79769e+308
1.79769e+3
08 0,00021683 0,00108079 yes
AT1G21550 1:7553100-7553876 WT MT OK 0 0,562868 1.79769e+308
1.79769e+3
08 0,00115582 0,00471497 yes
AT1G22120 1:7806308-7809632 WT MT OK 0 0,617354 1.79769e+308
1.79769e+3
08 2.48392e-06
1.91089e-
05 yes
AT1G31370 1:11238297-11239363 WT MT OK 0 1,46254 1.79769e+308
1.79769e+3
08 4.83523e-05
0,00028514
3 yes
APUM10 1:13253397-13255570 WT MT OK 0 0,581031 1.79769e+308
1.79769e+3
08 7.87855e-06
5.46603e-
05 yes
AT1G48700 1:18010728-18012871 WT MT OK 0 0,556525 1.79769e+308
1.79769e+3
08 6.53917e-05
0,00037473
6 yes
AT1G59077 1:21746209-21833195 WT MT OK 0 138,886 1.79769e+308
1.79769e+3
08 0,00122789 0,00496816 yes
AT1G60050 1:22121549-22123702 WT MT OK 0 0,370087 1.79769e+308
1.79769e+3
08 0,00117953 0,0048001 yes
Ddii et al., unpublished
AT4G15242 4:8705786-8706997 WT MT OK 0,00930712 17,9056 10,9098 -4,40523 1.05673e-05 7.13983e-05 yes
AT5G33251 5:12499071-12500433 WT MT OK 0,0498375 52,2837 10,0349 -9,8119 0 0 yes
AT4G12520 4:7421055-7421738 WT MT OK 0,0195111 15,8516 9,66612 -3,90043 9.60217e-05 0,000528904 yes
AT1G60020 1:22100651-22105276 WT MT OK 0,0118377 7,18823 9,24611 -7,50382 6.19504e-14 1.4988e-12 yes
AT5G15360 5:4987235-4989182 WT MT OK 0,0988273 56,4834 9,1587 -10,4392 0 0 yes
9
Plant Vascular Tissue
Development
□ Vascular tissue as a developmental model for GO analysis and MRN
modeling
Lehesranta etal., Trends in Plant Sci (2010)
10
WT hormonal mutant
Hormonal Control Over Vascular
Tissue Development
□ Plant Hormones Regulate Lignin Deposition in Plant Cell
Walls and Xylem Water Conductivity
WT mutant
lignified cell walls
Water Conductivity
WT hormonal mutants
11
WT hormonal mutant
Hormonal Control Over Vascular
Tissue Development
□ Transcriptional profiling via RNA sequencing
mRNA
Sequencing by Illumina and
number of transcripts determination
mRNA
cDNA cDNA
12
Results of –omics Studies vs
Biologically Relevant Conclusions
□ Transcriptional profiling yielded more then 9K differentially
regulated genes…
gene locus sample_1 sample_2 status value_1 value_2 log2(fold_change) test_stat p_value q_value significant
AT1G07795 1:2414285-2414967 WT MT OK 0 1,1804 1.79769e+308
1.79769e+3
08 6.88885e-05
0,00039180
1 yes
HRS1 1:4556891-4558708 WT MT OK 0 0,696583 1.79769e+308
1.79769e+3
08 6.61994e-06
4.67708e-
05 yes
ATMLO14 1:9227472-9232296 WT MT OK 0 0,514609 1.79769e+308
1.79769e+3
08 9.74219e-05
0,00053505
5 yes
NRT1.6 1:9400663-9403789 WT MT OK 0 0,877865 1.79769e+308
1.79769e+3
08 3.2692e-08
3.50131e-
07 yes
AT1G27570 1:9575425-9582376 WT MT OK 0 2,0829 1.79769e+308
1.79769e+3
08 9.76039e-06 6.647e-05 yes
AT1G60095 1:22159735-22162419 WT MT OK 0 0,688588 1.79769e+308
1.79769e+3
08 9.95901e-08
9.84992e-
07 yes
AT1G03020 1:698206-698515 WT MT OK 0 1,78859 1.79769e+308
1.79769e+3
08 0,00913915 0,0277958 yes
AT1G13609 1:4662720-4663471 WT MT OK 0 3,55814 1.79769e+308
1.79769e+3
08 0,00021683 0,00108079 yes
AT1G21550 1:7553100-7553876 WT MT OK 0 0,562868 1.79769e+308
1.79769e+3
08 0,00115582 0,00471497 yes
AT1G22120 1:7806308-7809632 WT MT OK 0 0,617354 1.79769e+308
1.79769e+3
08 2.48392e-06
1.91089e-
05 yes
AT1G31370 1:11238297-11239363 WT MT OK 0 1,46254 1.79769e+308
1.79769e+3
08 4.83523e-05
0,00028514
3 yes
APUM10 1:13253397-13255570 WT MT OK 0 0,581031 1.79769e+308
1.79769e+3
08 7.87855e-06
5.46603e-
05 yes
AT1G48700 1:18010728-18012871 WT MT OK 0 0,556525 1.79769e+308
1.79769e+3
08 6.53917e-05
0,00037473
6 yes
AT1G59077 1:21746209-21833195 WT MT OK 0 138,886 1.79769e+308
1.79769e+3
08 0,00122789 0,00496816 yes
AT1G60050 1:22121549-22123702 WT MT OK 0 0,370087 1.79769e+308
1.79769e+3
08 0,00117953 0,0048001 yes
Ddii et al., unpublished
AT4G15242 4:8705786-8706997 WT MT OK 0,00930712 17,9056 10,9098 -4,40523 1.05673e-05 7.13983e-05 yes
AT5G33251 5:12499071-12500433 WT MT OK 0,0498375 52,2837 10,0349 -9,8119 0 0 yes
AT4G12520 4:7421055-7421738 WT MT OK 0,0195111 15,8516 9,66612 -3,90043 9.60217e-05 0,000528904 yes
AT1G60020 1:22100651-22105276 WT MT OK 0,0118377 7,18823 9,24611 -7,50382 6.19504e-14 1.4988e-12 yes
AT5G15360 5:4987235-4989182 WT MT OK 0,0988273 56,4834 9,1587 -10,4392 0 0 yes
13
Gene Ontology Analysis
□ One of the possible approaches is to study gene ontology, i.e.
previously demonstrated association of genes to biological
processes
Ddii et al., unpublished
14
Gene Ontology Analysis
□ Several tools allow statistical evaluation of enrichment for
genes associated with specific processes
Eden et al., BMC Biinformatics (2009)
15
Gene Ontology Analysis
□ Several tools allow statistical evaluation of enrichment for
genes associated with specific processes
16
Gene Ontology Analysis
□ Several tools allow statistical evaluation of enrichment for
genes associated with specific processes
17
Gene Ontology Analysis
□ Several tools allow statistical evaluation of enrichment for
genes associated with specific processes
18
Gene Ontology Analysis
□ Several tools allow statistical evaluation of enrichment for
genes associated with specific processes
19
Gene Ontology Analysis
□ Several tools allow statistical evaluation of enrichment for
genes associated with specific processes
20
 Definition of Systems Biology
 Tools
 Gene Ontology analysis
 Bayesian Networks
Outline
21
Bayesian Networks
 What are Bayesian networks?
 Probabilistic Graphical Model that can be used to build models from
data and/or expert opinion
https://www.youtube.com/watch?v=4fcqyzVJwH
M
22
Bayesian Networks
 What are Bayesian Networks?
 Probabilistic Graphical Model that can be used to build models from
data and/or expert opinion
 can be used for a wide range of tasks including prediction, anomaly
detection, diagnostics, automated insight, reasoning, time series
prediction and decision making under uncertainty
 NODES
 each node represents a variable such as someone's height, age or gender.
A variable might be discrete, such as Gender = {Female, Male} or might be
continuous such as someone's age
 LINKS
 added between nodes to indicate that one node directly influences the
other
23
Bayesian Networks
NODES
LINKS/EDGES
24
Asia Bayesian Network
https://www.bayesserver.com/
25
 Definition of Systems Biology
 Tools
 Gene Ontology analysis
 Bayesian Networks
 Molecular/Gene Regulatory Networks Modeling
Outline
26
□ Vascular tissue as a developmental model for MRN modeling
Benitez and Hejatko, PLoS One, 2013
Molecular Regulatory
Networks Modeling
27
Molecular Regulatory Networks Modeling
□ Literature search for published data and creating small
database
Interaction Evidence References
A-ARRs –| CK
signaling
Double and higher order type-A ARR mutants show
increased sensitivity to CK.
Spatial patterns of A-type ARR gene expression and
CK response are consistent with partially redundant
function of these genes in CK signaling.
A-type ARRs decreases B-type ARR6-LUC.
Note: In certain contexts, however, some A-ARRs
appear to have effects antagonistic to other A-
ARRs.
[27]
[27]
[13]
[27]
AHP6 –| AHP ahp6 partially recovers the mutant phenotype of the
CK receptor WOL.
Using an in vitro phosphotransfer system, it was
shown that, unlike the AHPs, native AHP6 was
unable to accept a phosphoryl group. Nevertheless,
AHP6 is able to inhibit phosphotransfer from other
AHPs to ARRs.
[9]
[9]
Benitez and Hejatko, PLoS One, 2013
28
Molecular Regulatory Networks Modeling
□ Formulating logical rules defining the model dynamics
Network node Dynamical rule
CK 2 If ipt=1 and ckx=0
1 If ipt=1 and ckx=1
0 else
CKX 1 If barr>0 or arf=2
0 else
AHKs ahk=ck
AHPs 2 If ahk=2 and ahp6=0 and aarr=0
1 If ahk=2 and (ahp6+aarr<2)
1 If ahk=1 and ahp6<1
0 else
B-Type ARRs 1 If ahp>0
0 else
A-Type ARRs 1 If arf<2 and ahp>0
0 else
Benitez and Hejatko, PLoS One, 2013
29
Molecular Regulatory Networks Modeling
□ Specifying mobile elements and their model behaviour
30
Molecular Regulatory Networks Modeling
□ Preparing the first version of the model and its testing
31
Molecular Regulatory Networks Modeling
□ Specifying of missing interactions via informed predictions
Interaction Evidence References
A-ARRs –| CK
signaling
Double and higher order type-A ARR mutants show
increased sensitivity to CK.
Spatial patterns of A-type ARR gene expression and CK
response are consistent with partially redundant function of
these genes in CK signaling.
A-type ARRs decreases B-type ARR6-LUC.
Note: In certain contexts, however, some A-ARRs appear to
have effects antagonistic to other A-ARRs.
[27]
[27]
[13]
[27]
AHP6 –| AHP ahp6 partially recovers the mutant phenotype of the CK
receptor WOL.
Using an in vitro phosphotransfer system, it was shown that,
unlike the AHPs, native AHP6 was unable to accept a
phosphoryl group. Nevertheless, AHP6 is able to inhibit
phosphotransfer from other AHPs to ARRs.
[9]
[9]
CK → PIN7 radial
localization
Predicted interaction (could be direct or indirect)
Informed by the following data:
During the specification of root vascular cells in Arabidopsis
thaliana, CK regulates the radial localization of PIN7.
Expression of PIN7:GFP and PIN7::GUS is upregulated by
CK with no significant influence of ethylene.
In the root, CK signaling is required for the CK regulation of
PIN1, PIN3, and PIN7. Their expression is altered in wol,
cre1, ahk3 and ahp6 mutants.
[18]
[18,20]
[19]
CK→ APL Predicted interaction (could be direct or indirect)
Consistent with the fact that APL overexpression prevents or
delays xylem cell differentiation, as does CKs.
Partially supported by microarray data and phloem-specific
expression patterns of CK response factors.
[21]
(TAIR,
ExpressionSet:
1005823559,
[22])
32
Molecular Regulatory Networks Modeling
□ Preparing the next version of the model and its testing
Benitez and Hejatko, PLoS One, 2013
33
□ Good model should be able to simulate reality
Benitez and Hejatko, PLoS One, 2013
Molecular Regulatory
Networks Modeling
34
□ Formulating equations describing the relationships in the model
Molecular Regulatory
Networks Modeling
Static nodes: gn(t+1)=Fn(gn1(t),gn2(t),..., gnk(t))
Mobile nodes: g(t+1)T [i]= H(g(t) [i]+ D (g(t) [i+1]+g(t) [i-1] – N(g(t) [i]))-b)
state in the time t+1
state in the time tlogical rule function
state in the time t+1
Amount if TDIF or MIR165 in cell i
proportion of movable element
constant corresponding to a degradation term
35
□ Good model should be able to simulate reality
Benitez and Hejatko, submitted
Molecular Regulatory
Networks Modeling
Static nodes: gn(t+1)=Fn(gn1(t),gn2(t),..., gnk(t))
Mobile nodes: g(t+1)T [i]= H(g(t) [i]+ D (g(t) [i+1]+g(t) [i-1] – N(g(t) [i]))-b)
36
Molecular Regulatory Networks Modeling
Benitez and Hejatko, submitted
□ The good model should be able to simulate reality
37
Molecular Regulatory Networks Modeling
Benitez and Hejatko, submitted
□ Simulation of mutants
38
 Definition of Systems Biology
 Tools
 Gene Ontology analysis
 Bayesian Networks
 Molecular/Gene Regulatory Networks Modeling
 Inferring Gene Regulatory Networks from Large Omics
Datasets
Outline
39
Systems Biology in
Cancer Research
40
miRNA/mRNA Profiling
Guo et al., Mol Med Reports, 2017
41 Benkova and Hejatko,Plant Mol Biol (2008)
proximalroot
meristem
distalroot
meristem
Inferring Gene
Regulatory Networks
42
Klidové centrum
Quiescent centre
Kolumela
Columella cell files
Iniciály kolumely
Columella initials
Epiderims
Epidermis
Kortex
Cortex
Endodermis
Endodermis
Iniciály stéle
Stele initials
proximalroot
meristem
distalroot
meristem
Postranní kořenová čepička
Lateral root cap
Iniciály epidermis
Epidermis initials
Iniciála endodermis a kortexu
Endodermis and cortex initial
Gene Regulatory
Networks
43 Birnbaum et al., Science, 2003
Gene Regulatory
Networks - GENIST
de Luis Balaguer et al., PNAS, 2017
 Inferring GRNs via GENIST
 GEne regulatory Network
Inference from SpatioTemporal
data algorithm
 Combining spatial- and timespecific
gene expression
profiles
44
Combining Large Omics
Datasets GENES
TISSUE/TIME
45
Gene Regulatory
Networks - GENIST
 Inferring GRNs via GENIST
 Clustering of genes
 Expression similarity under
various conditions/genetic
backgrounds, time points, …
 Inferring intra-cluster
connections
 Selection of potential
regulators and co-
regulators
 Based on the time
correlation in the change
of expression and/or user
specification
 Dynamic Bayesian
Network modeling
Haeseleer, Computational Biology, 2005
46
 Inferring GRNs via GENIST
 Clustering of genes
 Expression similarity under
various conditions/genetic
backgrounds, time points, …
 Inferring intra-cluster
connections
 Selection of potential
regulators and co-
regulators
 Based on the time
correlation in the change
of expression and/or user
specification
 Dynamic Bayesian
Network modeling
de Luis Balaguer et al., PNAS, 2017
Gene Regulatory
Networks - GENIST
47
Gene Regulatory
Networks - GENIST
de Luis Balaguer et al., PNAS, 2017
48 de Luis Balaguer et al., PNAS, 2017
Gene Regulatory
Networks - GENIST
49
Indirect PAN targets
feed-back loops
Gene Regulatory
Networks - GENISTMODEL PREDICTION
EXPERIMENTAL
VERIFICATION
50
 Definition of Systems Biology
 Tools
 Gene Ontology analysis
 Bayesian Networks
 Molecular/Gene Regulatory Networks Modeling
 Inferring Gene Regulatory Networks from Large Omics
Datasets
Summary
51
Discussion