Theory and Limitations of Genetic Network Inference from Microarray Data Export

@article{citeulike:1939672, abstract = {Since the advent of gene expression microarray technology more than 10 years ago, many computational approaches have been developed aimed at using statistical associations between mRNA abundance profiles to predict transcriptional regulatory interactions. The ultimate goal is to develop causal network models describing the transcriptional influences that genes exert on each other (via their protein products), which can be used to predict network disruptions (e.g., mutations) leading to a disease phenotype, as well as the appropriate therapeutic intervention. However, microarray data measure only a small component of the interacting variables in a genetic regulatory network, as cells are known to regulate gene expression via many diverse mechanisms. Although many researchers have acknowledged the questionable interpretation of statistical dependencies between mRNA profiles, very little work has been done on theoretically characterizing the nature of inferred dependencies using models that account for unobserved interacting variables. In this work, we review the theory behind reverse engineering algorithms derived from three separate disciplines2014system control theory, graphical models, and information theory2014and highlight several mathematical relationships between the various methods. We then apply recent theoretical work on constructing graphical models with latent variables to the context of reverse engineering genetic networks. We demonstrate that even the addition of simple latent variables induces statistical dependencies between non2013directly interacting (e.g., co-regulated) genes that cannot be eliminated by conditioning on any observed variables.}, address = {Department of Biomedical Informatics, Columbia University, New York, New York, USA}, author = {Margolin, Adam A. and Andrea Califan, O.}, citeulike-article-id = {1939672}, citeulike-linkout-0 = {http://www.blackwell-synergy.com/doi/abs/10.1196/annals.1407.019}, citeulike-linkout-1 = {http://dx.doi.org/10.1196/annals.1407.019}, citeulike-linkout-2 = {http://www.ingentaconnect.com/content/bsc/nyas/2007/00001115/00000001/art00005}, citeulike-linkout-3 = {http://view.ncbi.nlm.nih.gov/pubmed/17925348}, citeulike-linkout-4 = {http://www.hubmed.org/display.cgi?uids=17925348}, citeulike-linkout-5 = {http://www3.interscience.wiley.com/cgi-bin/abstract/117986064/ABSTRACT}, doi = {10.1196/annals.1407.019}, issn = {1749-6632}, journal = {Annals of the New York Academy of Sciences}, keywords = {dependence, gene-expression, genetic, inference, latent, microarray, network, reverse-engineering, review, statistical, theory}, month = {December}, number = {Reverse Engineering Biological Networks: Opportunities and Challenges in Computational Methods for Pathway Inference}, pages = {51--72}, posted-at = {2009-09-30 18:01:54}, priority = {2}, publisher = {Blackwell Publishing}, title = {Theory and Limitations of Genetic Network Inference from Microarray Data}, url = {http://dx.doi.org/10.1196/annals.1407.019}, volume = {1115}, year = {2007} }

TY - JOUR ID - citeulike:1939672 L3 - citeulike-article-id:1939672 N2 - Since the advent of gene expression microarray technology more than 10 years ago, many computational approaches have been developed aimed at using statistical associations between mRNA abundance profiles to predict transcriptional regulatory interactions. The ultimate goal is to develop causal network models describing the transcriptional influences that genes exert on each other (via their protein products), which can be used to predict network disruptions (e.g., mutations) leading to a disease phenotype, as well as the appropriate therapeutic intervention. However, microarray data measure only a small component of the interacting variables in a genetic regulatory network, as cells are known to regulate gene expression via many diverse mechanisms. Although many researchers have acknowledged the questionable interpretation of statistical dependencies between mRNA profiles, very little work has been done on theoretically characterizing the nature of inferred dependencies using models that account for unobserved interacting variables. In this work, we review the theory behind reverse engineering algorithms derived from three separate disciplines2014system control theory, graphical models, and information theory2014and highlight several mathematical relationships between the various methods. We then apply recent theoretical work on constructing graphical models with latent variables to the context of reverse engineering genetic networks. We demonstrate that even the addition of simple latent variables induces statistical dependencies between non2013directly interacting (e.g., co-regulated) genes that cannot be eliminated by conditioning on any observed variables. IS - Reverse Engineering Biological Networks: Opportunities and Challenges in Computational Methods for Pathway Inference JF - Annals of the New York Academy of Sciences SN - 1749-6632 AD - Department of Biomedical Informatics, Columbia University, New York, New York, USA EP - 72 TI - Theory and Limitations of Genetic Network Inference from Microarray Data PB - Blackwell Publishing VL - 1115 SP - 51 KW - dependence KW - gene-expression KW - genetic KW - inference KW - latent KW - microarray KW - network KW - reverse-engineering KW - review KW - statistical KW - theory AU - Margolin, Adam AU - Andrea Califan, O PY - 2007/12// UR - http://dx.doi.org/10.1196/annals.1407.019 ER -