1. Drivers, Passengers, and Biomarkers via Network Enrichment Analysis of Tumor Molecular Profiles
Andrey Alexeyenko

2. FunCoup: on-line interactome resource
Andrey Alexeyenko and Erik L.L. Sonnhammer (2009) Global networks of functional coupling in eukaryotes from comprehensive data integration. Genome Research.

3. Do gene networks tell any story?
State-of-the-art method to beat:
Frequency analysis of somatic mutations
Do gene networks tell any story?
Yellow diamonds: somatic mutations in prostate cancer
Pink crosses: also mutated in glioblastome (TCGA)

4. Network analysis State-of-the-art:
The global gene/protein networks are widely available
They do contain biological signal
It is possible to use them in biological research
Network visualization is just great!
Is used for (or not so much so far…):
Exploratory (de novo) analysis
Functional annotation
Hypothesis formulation
Hypothesis testing, significance evaluation
Network-based data transformation and processing

5. Somatic mutations: drivers vs
Somatic mutations: drivers vs. passengers Our key assumption is: drivers within one tumor genome should be functionally related to each other and/or to known cancer pathways
data from The Cancer Genome Atlas

6. Network enrichment analysis How to prove significance
Network enrichment analysis How to prove significance? compare to a reference and quantify
In the actual network
In the randomized network
Quantification:
N links_real = 12
N links_expected = 4.65
Standard deviation = 1.84
Question:
Is ANXA2 related to TGFbeta signaling?
Z = (N links_observed – N links_expected) / SD = 3.98
P-value =
FDR < 0.1

7. Could the emerged mutations interact with each other?
Could a given mutation interact
with the others? PM
Point
mutation
driver?
CNA CP
CNA driver? or C A D B
Could the emerged
gene copy number change
interact with the point mutations?
Could the emerged
mutations (either point or copy number change) interact with known cancer pathways?
Simon Kebede Merid, Daria Goranskaya, Andrey Alexeyenko Distinguishing between driver and passenger mutations in individual cancer genomes by network enrichment analysis BMC Bioinformatics, 2014, 15:308.

8. Is NTRK1 a driver in the GBM tumor TCGA-02-0014?
data from
The Cancer Genome Atlas

9. Validation of candidate disease genes (work with Jonathan Prince, MEB, Karolinska Inst.)
Genetic association of sequence variants near AGER/NOTCH4 and dementia.
Bennet AM, Reynolds CA, Eriksson UK, Hong MG, Blennow K, Gatz M, Alexeyenko A, Pedersen NL, Prince JA.
J Alzheimers Dis. 2011;24(3):
Genome-wide pathway analysis implicates intracellular transmembrane protein transport in Alzheimer disease.
Hong MG, Alexeyenko A, Lambert JC, Amouyel P, Prince JA.
J Hum Genet Oct;55(10): Epub 2010 Jul 29.
Analysis of lipid pathway genes indicates association of sequence variation near SREBF1/TOM1L2/ATPAF2 with dementia risk.
Reynolds CA, Hong MG, Eriksson UK, Blennow K, Wiklund F, Johansson B, Malmberg B, Berg S, Alexeyenko A, Grönberg H, Gatz M, Pedersen NL, Prince JA.
Hum Mol Genet May 15;19(10): Epub 2010 Feb 18.
Question: Is there extra evidence for GWAS-candidates to be involved?
Answer: Yes, for some…

10. Genes altered by RhoA knock-out in fibroblasts, before and after confrontation with PC3 cells

11. Genes altered by RhoA knock-out in fibroblasts, before and after confrontation with PC3 cells

13. Poor overlap between gene signatures of relapse
Usually the overlap between signatures is negligible because of e.g.:
Different sub-types of patient population,
Different microarray platforms.
However the main, “unavoidable”
reasons are:
Cancer heterogeneity
Dimensionality curse
From Roepman et al., 2007

14. “CCLE”:
>500 lines X 24 drugs
“CTD2”:
~242 lines X 354 drugs
“CGP”:
~500 lines X 130 drugs
Poor overlap between drug sensitivity profiles

16. Functional characterization of novel gene sets
Gene set enrichment analysis
Network enrichment analysis
Altered genes
Functional set ?
Alexeyenko A, Lee W, … Pawitan Y. Network enrichment analysis: extension of gene-set enrichment analysis to gene networks. BMC Bioinformatics, 2012
Huang da W, … Lempicki RA. Systematic and integrative analysis of large gene lists using DAVID bioinformatics resources. Nature Protocols, 2009

17. Network enrichment analysis How to prove significance
Network enrichment analysis How to prove significance? compare to a reference and quantify
In the actual network
In the randomized network
Quantification:
N links_real = 12
N links_expected = 4.65
Standard deviation = 1.84
Question:
Is ANXA2 related to TGFbeta signaling?
Z = (N links_observed – N links_expected) / SD = 3.98
P-value =
FDR < 0.1

18. NEA: molecular profiles rendered into pathway space
P53 signaling
Sample B
Sample A
Sample D
Apoptosis
Sample C
Sample E
Cell cycle

19. Network enrichment analysis
Altered gene sets
Global network
1) Text box:
i, j, k, l, m, n i
or 2) uploaded gene lists:
FGS i k l m k l m n o
AGS i k l m
or 3) Venn diagram from uploaded
DE values: k m n i K j
Network enrichment analysis
Functional gene sets, FGS
Ashwini Jeggari, Zhanna Alekseenko, José Dias, Johan Ericson, Andrey Alexeyenko
EviNet: a web platform for network enrichment analysis with flexible definition of gene sets, under review.

20. Same drug-pathway associations can be traced from in vitro to in vivo
”drug X feature“ association was proven given covariates:
Age
Race
Gender
Tumor stage
Year of diagnosis
Same drug-pathway associations
can be traced from in vitro to in vivo
Example 3
Both

21. A 5-pathway model of sensitivity to Prima-1-met in cancer cell lines remains valid in independent validation by ACT cell line screen (KI)
Marcela Franco, Ashwini Jeggari, Galina Selivanova, Andrey Alexeyenko Robust prediction of sensitivity to anti-cancer drugs using network-based scores, manuscript

23. cell line-specific gene lists
P-values of correlation between cell line sensitivity to anti-cancer drug Atra and the cell line transcriptome
Log10(p) values are plotted for the two drug screens, identified by the first author
NEA pathway scores for
cell line-specific gene lists
Gene expression profiles

24. https://research.scilifelab.se/andrej_alexeyenko/HyperSet/

27. Network enrichment analysis: what software to use?
Perl software NEA.pl (focus on multi-functionality):
Simon Kebede Merid, Daria Goranskaya, Andrey Alexeyenko Distinguishing between driver and passenger mutations in individual cancer genomes by network enrichment analysis BMC Bioinformatics, 2014, 15:308.
C++ “crosstalk” tool (focus on network permutation):
McCormack T, Frings O, Alexeyenko A, Sonnhammer EL. Statistical assessment of crosstalk enrichment between gene groups in biological networks. PLoS One. 2013;8(1):e54945.
R package NEA (outdated):
Alexeyenko A, Lee W, … Pawitan P (2012). Network enrichment analysis: extension of gene-set enrichment analysis to gene networks. BMC Bioinformatics
R package NEArender (focus on quick matrix conversion “raw data -> NEA scores”):
to be published
Follow us at

28. Acknowledgements Simon Merid Ashwini Jeggari Darya Goranskaya Pan Lu
Erik Sonnhammer
Martin Klammer
Sanjit Roopra
Ted McCormack
Oliver Frings
Jonathan Prince
Yudi Pawitan
Setia Pramana
WooJoo Lee
Joakim Lundeberg
Pelin Akan
Per Kraulis
Ingemar Ernberg
Simon Merid
Ashwini Jeggari
Darya Goranskaya
Pan Lu

29. Alexeyenko A, Wassenberg DM, Lobenhofer EK, Yen J, Linney E, Sonnhammer EL, Meyer JN. Dynamic zebrafish interactome reveals transcriptional mechanisms of dioxin toxicity. PLoS One May 5;5(5):e10465.
Alexeyenko A, Lee W, Pernemalm M, Guegan J, Dessen P, Lazar V, Lehtiö J, Pawitan Y. Network enrichment analysis: extension of gene-set enrichment analysis to gene networks. BMC Bioinformatics Sep 11;13:226.
McCormack T, Frings O, Alexeyenko A, Sonnhammer EL. Statistical assessment of crosstalk enrichment between gene groups in biological networks. PLoS One. 2013;8(1):e54945.
Frings O, Alexeyenko A, Sonnhammer EL. MGclus: network clustering employing shared neighbors. Mol Biosyst Jul;9(7):1670-5
Merid SK, Goranskaya D, Alexeyenko A Distinguishing between driver and passenger mutations in individual cancer genomes by network enrichment analysis BMC Bioinformatics, 2014, 15:308.