Vol. 28 no. 5 2012, pages 694–700
Modeling community-wide molecular networks of multicellular
Divisions of Experimental Hematology and Cancer Biology, Human Genetics and Biomedical Informatics, Cincinnati
Children’s Hospital Medical Center, Cincinnati, OH, USA
Associate Editor: Trey Ideker
Advance Access publication December 30, 2011
Motivation: Multicellular systems, such as tissues, are composed
of different cell types that form a heterogeneous community.
Behavior of these systems is determined by complex regulatory
networks within (intracellular networks) and between (intercellular
networks) cells. Increasingly more studies are applying genome-
wide experimental approaches to delineate the contributions of
individual cell types (e.g. stromal, epithelial, vascular cells) to
collective behavior of heterogeneous cell communities (e.g. tumors).
Although many computational methods have been developed for
analyses of intracellular networks based on genome-scale data,
these efforts have not been extended toward analyzing genomic data
from heterogeneous cell communities.
Results: Here, we propose a network-based approach for analyses
of genome-scale data from multiple cell types to extract community-
wide molecular networks comprised of intra- and intercellular
interactions. Intercellular interactions in this model can be physical
interactions between proteins or indirect interactions mediated by
secreted metabolites of neighboring cells. Applying this method on
data from a recent study on xenograft mouse models of human
lung adenocarcinoma, we uncover an extensive network of intra-
and intercellular interactions involved in the acquired resistance to
Supplementary information: Supplementary data are available at
Received on September 15, 2011; revised on November 10, 2011;
accepted on December 25, 2011
Individual cells within a heterogeneous cell population interact with
each other through secreted molecules and membrane proteins,
sometimes referred to as cross-talk (Frankenstein et al., 2006;
Jahoda and Christiano, 2011; Kiel and Morrison, 2008). At the
molecular level, this population can be viewed as a community-
wide network of molecular interactions comprising intracellular
interactions within each cell as well as intercellular interactions
of molecules of different cells. Since population characteristics
as a whole are highly dependent on the intra- as well as
intercellular networks, the global architecture of the community-
wide molecular network (CMN, made of intra- and intercellular
∗To whom correspondence should be addressed.
molecular interactions) can determine the collective behavior of
heterogeneous cell communities.
Network-based analyses of genomic data have shed light on the
global organization of intracellular networks contributing to normal
and malignant behavior of cells (Basso et al., 2005; Calvano et al.,
2005; Chuang et al., 2007;Tomlins et al., 2007). Mounting evidence
now suggests that interplays of cells within a microenvironment
can give rise to complex population behavior (Frankenstein et al.,
2006; Jahoda and Christiano, 2011). Such complex interactions of
cells within a population have been observed in developmental
processes (Kirouac et al., 2010; Lai, 2004), in stem cell niches
microenvironments (Boersma et al., 2008; Coussens and Werb,
2002). However, most of the studies on deciphering the complex
pattern of molecular network interactions in such multicellular
systems and their role in population-wide collective behavior have
been focused on a limited number of molecules. Although some
notable large-scale studies in some experimental systems have
been undertaken (Frankenstein et al., 2006; Kirouac et al., 2010),
the computational methodology of analysis primarily involved
candidate-based approaches, limiting the scope of analysis. More
powerful computational methods for analyses of genomic data from
heterogeneous cell populations would therefore greatly enhance our
ability to gain insight into the organization of CMNs and their role
in the collective population behavior.
Here, in order to enable modeling of molecular networks
of whole-cell populations, we developed a network model of
community-wide molecular interactions by combining intracellular
interactions from each cell type and their intercellular connections
into a single global network (community molecular network)
(Fig. 1). We use this global network in conjunction with the
genome-wide gene expression data from different cell types to
extract networks of interest showing community-wide molecular
interactions most highlighted by the data. We integrate genomic
data with the global network using NetWalk (Komurov et al.,
2010), a computational algorithm for seamless integration of
genomic data with molecular networks. The advantage of NetWalk
compared to other network analysis tools is that NetWalk takes
into account the whole-data distribution without requiring statistical
cutoffs or predetermined gene lists of interest. NetWalk output
is a distribution of Edge Flux (EF) values containing numeric
score of relevance assigned to each interaction in the network.
EF values, just like in gene expression data, can be used for
further statistical analyses, allowing for direct network-based
statistical analyses. Using this platform, we present an analysis
of recently published gene expression data from epithelial and
© The Author 2011. Published by Oxford University Press. All rights reserved. For Permissions, please email: firstname.lastname@example.org
Cascone,T. et al. (2011) Upregulated stromal EGFR and vascular remodeling in mouse
xenograft models of angiogenesis inhibitor-resistant human lung adenocarcinoma.
J. Clin. Invest., 121, 1313–1328.
Cerami,E.G. et al. (2011) Pathway Commons, a web resource for biological pathway
data. Nucleic Acids Res., 39, D685–D690.
Chatr-aryamontri,A. et al. (2007) MINT: the Molecular INTeraction database. Nucleic
Acids Res., 35, D572–D574.
Chuang,H.Y. et al. (2007) Network-based classification of breast cancer metastasis.
Mol. Syst. Biol., 3, 140.
Coussens,L.M. and Werb,Z. (2002) Inflammation and cancer. Nature, 420, 860–867.
Dienel,G.A. and Cruz,N.F. (2003) Neighborly interactions of metabolically-activated
astrocytes in vivo. Neurochem. Int., 43, 339–354.
Frankenstein,Z. et al. (2006) The immune-body cytokine network defines a social
architecture of cell interactions. Biol. Direct., 1, 32.
Jahoda,C.A. and Christiano,A.M. (2011) Niche crosstalk: intercellular signals at the
hair follicle. Cell, 146, 678–681.
Joshi-Tope,G. et al. (2005) Reactome: a knowledgebase of biological pathways. Nucleic
Acids Res., 33, D428–D432.
Kanehisa,M. and Goto,S. (2000) KEGG: kyoto encyclopedia of genes and genomes.
Nucleic Acids Res., 28, 27–30.
Kerrien,S. et al. (2007) IntAct–open source resource for molecular interaction data.
Nucleic Acids Res., 35, D561–D565.
Kiel,M.J. and Morrison,S.J. (2008) Uncertainty in the niches that maintain
haematopoietic stem cells. Nat. Rev. Immunol., 8, 290–301.
Kirouac,D.C. et al. (2010) Dynamic interaction networks in a hierarchically organized
tissue. Mol. Syst. Biol., 6, 417.
Komurov,K. et al. (2010) use of data-biased random walks on graphs for the retrieval
of context-specific networks from genomic data. PLoS Comput. Biol., 6, e1000889.
and stromal cells in colorectal carcinomas: a metabolic survival role for tumor-
associated stroma. Cancer Res., 66, 632–637.
Lai,E.C. (2004) Notch signaling: control of cell communication and cell fate.
Development, 131, 965–973.
Lewis,N.E. et al. (2010) Large-scale in silico modeling of metabolic interactions
between cell types in the human brain. Nat. Biotechnol., 28, 1279–1285.
Res., 34, D411–D414.
Pavlides,S. et al. (2009) The reverse Warburg effect: aerobic glycolysis in cancer
associated fibroblasts and the tumor stroma. Cell Cycle, 8, 3984–4001.
Sonveaux,P. et al. (2008) Targeting lactate-fueled respiration selectively kills hypoxic
tumor cells in mice. J. Clin. Invest., 118, 3930–3942.
Stark,C. et al. (2006) BioGRID: a general repository for interaction datasets. Nucleic
Acids Res., 34, D535–D539.
Tomlins,S.A. et al. (2007) Integrative molecular concept modeling of prostate cancer
progression Nat. Genet., 39, 41–51.
Wingender,E. et al. (2000) TRANSFAC: an integrated system for gene expression
regulation Nucleic Acids Res., 28, 316–319.
Acids Res., 37, D603–D610.