Detecting subnetwork-level dynamic correlations

Motivation: The biological regulatory system is highly dynamic. The correlations between many functionally related genes change over different biological conditions. Finding dynamic relations on the existing biological network may reveal important regulatory mechanisms. Currently no method is available to detect subnetwork-level dynamic correlations systematically on the genome-scale network. Two major issues hampered the development. The first is gene expression profiling data usually do not contain time course measurements to facilitate the analysis of dynamic relations, which can be partially addressed by using certain genes as indicators of biological conditions. Secondly, it is unclear how to effectively delineate subnetworks, and define dynamic relations between them. Results: Here we propose a new method named LANDD (Liquid Association for Network Dynamics Detection) to find subnetworks that show substantial dynamic correlations, as defined by subnetwork A is concentrated with Liquid Association scouting genes for subnetwork B. The method produces easily interpretable results because of its focus on subnetworks that tend to comprise functionally related genes. Also, the collective behaviour of genes in a subnetwork is a much more reliable indicator of underlying biological conditions compared to using single genes as indicators. We conducted extensive simulations to validate the method’s ability to detect subnetwork-level dynamic correlations. Using a real gene expression dataset and the human protein-protein interaction network, we demonstrate the method links subnetworks of distinct biological processes, with both confirmed relations and plausible new functional implications. We also found signal transduction pathways tend to show extensive dynamic relations with other functional groups. Availability and Implementation: The R package is available at https://cran.r-project.org/web/packages/LANDD. Contacts: yunba@pcom.edu, jwlu33@hotmail.com or tianwei.yu@emory.edu Supplementary information: Supplementary data are available at Bioinformatics online.

[1]  Daniel J. Brass,et al.  Network Analysis in the Social Sciences , 2009, Science.

[2]  Wei Pan,et al.  Network‐based genomic discovery: application and comparison of Markov random‐field models , 2010, Journal of the Royal Statistical Society. Series C, Applied statistics.

[3]  C. Sander,et al.  Mutual exclusivity analysis identifies oncogenic network modules. , 2012, Genome research.

[4]  Guido Sanguinetti,et al.  MMG: a probabilistic tool to identify submodules of metabolic pathways , 2008, Bioinform..

[5]  Jingkai Yu,et al.  Mining breast cancer genes with a network based noise-tolerant approach , 2013, BMC Systems Biology.

[6]  David Warde-Farley,et al.  Dynamic modularity in protein interaction networks predicts breast cancer outcome , 2009, Nature Biotechnology.

[7]  A. Barabasi,et al.  Network medicine--from obesity to the "diseasome". , 2007, The New England journal of medicine.

[8]  Ram Rup Sarkar,et al.  Comparison of human cell signaling pathway databases—evolution, drawbacks and challenges , 2015, Database J. Biol. Databases Curation.

[9]  M. Gerstein,et al.  Genomic analysis of regulatory network dynamics reveals large topological changes , 2004, Nature.

[10]  J. Byrd,et al.  Antagonizing ClpP: A New Power Play in Targeted Therapy for AML. , 2015, Cancer cell.

[11]  J. Loscalzo,et al.  The Emerging Paradigm of Network Medicine in the Study of Human Disease , 2012, Circulation research.

[12]  M. Marino,et al.  Role of tyrosine kinase signaling in estrogen-induced LDL receptor gene expression in HepG2 cells. , 2002, Biochimica et biophysica acta.

[13]  Albert-László Barabási,et al.  Universality in network dynamics , 2013, Nature Physics.

[14]  Yibo Wu,et al.  GOSemSim: an R package for measuring semantic similarity among GO terms and gene products , 2010, Bioinform..

[15]  Guido Sanguinetti,et al.  Hybrid regulatory models: a statistically tractable approach to model regulatory network dynamics , 2013, Bioinform..

[16]  Ker-Chau Li,et al.  A system for enhancing genome-wide coexpression dynamics study. , 2004, Proceedings of the National Academy of Sciences of the United States of America.

[17]  Jonathan C. Cohen,et al.  Molecular mechanisms of autosomal recessive hypercholesterolemia , 2003, Current opinion in lipidology.

[18]  Haiyuan Yu,et al.  HINT: High-quality protein interactomes and their applications in understanding human disease , 2012, BMC Systems Biology.

[19]  Robert Gentleman,et al.  Using GOstats to test gene lists for GO term association , 2007, Bioinform..

[20]  D. Zajonc,et al.  Atomic structure of the autosomal recessive hypercholesterolemia phosphotyrosine-binding domain in complex with the LDL-receptor tail , 2012, Proceedings of the National Academy of Sciences.

[21]  A. Barabasi,et al.  Network medicine : a network-based approach to human disease , 2010 .

[22]  Albert-László Barabási,et al.  Erratum: Universality in network dynamics , 2013 .

[23]  Wei Pan,et al.  Bayesian Joint Modeling of Multiple Gene Networks and Diverse Genomic Data to Identify Target Genes of a Transcription Factor. , 2012, The annals of applied statistics.

[24]  T. Ideker,et al.  Differential network biology , 2012, Molecular systems biology.

[25]  Tianwei Yu,et al.  A BAYESIAN NONPARAMETRIC MIXTURE MODEL FOR SELECTING GENES AND GENE SUBNETWORKS. , 2014, The annals of applied statistics.

[26]  Edward R. Dougherty,et al.  Identification of diagnostic subnetwork markers for cancer in human protein-protein interaction network , 2010, BMC Bioinformatics.

[27]  Ker-Chau Li,et al.  Biomarkers and transcriptome profiling of lung cancer , 2012, Respirology.

[28]  Damian Szklarczyk,et al.  Protein-protein interaction databases. , 2015, Methods in molecular biology.

[29]  Zhaohui S. Qin,et al.  EgoNet: identification of human disease ego-network modules , 2014, BMC Genomics.

[30]  Gary D Bader,et al.  Inhibition of the Mitochondrial Protease ClpP as a Therapeutic Strategy for Human Acute Myeloid Leukemia. , 2015, Cancer cell.

[31]  Jonathan C. Cohen,et al.  ARH Is a Modular Adaptor Protein That Interacts with the LDL Receptor, Clathrin, and AP-2* , 2002, The Journal of Biological Chemistry.

[32]  Leena Peltonen,et al.  Finding disease candidate genes by liquid association , 2007, Genome Biology.

[33]  Vwani P. Roychowdhury,et al.  An Information Theoretic Exploratory Method for Learning Patterns of Conditional Gene Coexpression from Microarray Data , 2008, IEEE/ACM Transactions on Computational Biology and Bioinformatics.

[34]  Albert,et al.  Emergence of scaling in random networks , 1999, Science.

[35]  D. Mevorach,et al.  ‘Danger’ effect of low‐density lipoprotein (LDL) and oxidized LDL on human immature dendritic cells , 2007, Clinical and experimental immunology.

[36]  Korbinian Strimmer,et al.  A unified approach to false discovery rate estimation , 2008, BMC Bioinformatics.

[37]  Ker-Chau Li,et al.  Genome-wide coexpression dynamics: Theory and application , 2002, Proceedings of the National Academy of Sciences of the United States of America.

[38]  R. Tibshirani,et al.  Empirical bayes methods and false discovery rates for microarrays , 2002, Genetic epidemiology.

[39]  V.P. Roychowdhury,et al.  An Information Theoretic Exploratory Method for Learning Patterns of Conditional Gene Coexpression from Microarray Data , 2008, IEEE/ACM Transactions on Computational Biology and Bioinformatics.

[40]  K. Chang,et al.  Aberrant expression of CD7 in myeloblasts is highly associated with de novo acute myeloid leukemias with FLT3/ITD mutation. , 2008, American journal of clinical pathology.

[41]  Robert Clarke,et al.  Identifying protein interaction subnetworks by a bagging Markov random field-based method , 2012, Nucleic acids research.

[42]  Hongzhe Li,et al.  A penalized likelihood approach for bivariate conditional normal models for dynamic co-expression analysis. , 2011, Biometrics.

[43]  G. Paulsson-Berne,et al.  Inhibition of T cell response to native low-density lipoprotein reduces atherosclerosis , 2010, The Journal of experimental medicine.

[44]  Jonathan C. Cohen,et al.  The Modular Adaptor Protein Autosomal Recessive Hypercholesterolemia (ARH) Promotes Low Density Lipoprotein Receptor Clustering into Clathrin-coated Pits* , 2005, Journal of Biological Chemistry.

[45]  Matthieu Latapy,et al.  Computing Communities in Large Networks Using Random Walks , 2004, J. Graph Algorithms Appl..

[46]  Cheng Cheng,et al.  In Vivo Response to Methotrexate Forecasts Outcome of Acute Lymphoblastic Leukemia and Has a Distinct Gene Expression Profile , 2008, PLoS medicine.