## Design of probabilistic Boolean networks under the requirement of contextual data consistency (2006)

Venue: | IEEE Trans. Signal Process |

Citations: | 5 - 4 self |

### BibTeX

@ARTICLE{Dougherty06designof,

author = {Edward R. Dougherty and Yufei Xiao and Student Member},

title = {Design of probabilistic Boolean networks under the requirement of contextual data consistency},

journal = {IEEE Trans. Signal Process},

year = {2006},

volume = {54},

pages = {3603--3613}

}

### OpenURL

### Abstract

Abstract—A key issue of genomic signal processing is the design of gene regulatory networks. A probabilistic Boolean network (PBN) is composed of a family of Boolean networks. It stochastically switches between its constituent networks (contexts). For network design, connectivity and transition rules must be inferred from data via some optimization criterion. Except rarely, the optimal rule for a gene will not be a perfect predictor because there will be inconsistencies in the data. It would be natural to model these inconsistencies to reflect changes in PBN contexts. If we assume inconsistencies result from the data arising from a random function, then design involves finding the realizations of a random function and the probability mass on those realizations so that the resulting random function best fits the data relative to the expectation of its output and does so using a minimal number of realizations. We propose PBN design satisfying the biological assumption that data are consistent within a context, for which the distribution of the network agrees with the empirical distribution of the data, and such that this is accomplished with a minimal number of contexts. The design also satisfies the biological constraint that, because the network spends the great majority of time in its attractors, all data states should be attractor states in the model. Index Terms—Data consistency, gene regulatory network, graphical model, network inference. I.

### Citations

1013 |
Origins of order: Self-Organization and selection in evolution
- Kauffman
- 1993
(Show Context)
Citation Context ...for the network. Nonattractor states are transient. They are visited at most once on any network trajectory. When modeling genetic regulatory networks, attractors are often identified with phenotypes =-=[8]-=-. Real biological systems are typically assumed to have short attractor cycles, with singleton attractors being of special import. Much of our interest concerns singleton attractors. By definition, th... |

236 | Probabilistic Boolean Networks: A Rule-Based Uncertainty Model for Gene Regulatory Networks
- Shmulevich, Gluhovsky, et al.
- 2003
(Show Context)
Citation Context ...TION PROBABILISTIC Boolean networks (PBNs) represent an interface between the determinism of Boolean networks and the probabilistic nature of Bayesian networks by incorporating rule-based uncertainty =-=[15]-=-. This compromise is important because rule-based dependencies between genes are biologically meaningful, while mechanisms for handling uncertainty are conceptually and empirically necessary. The bina... |

183 |
Molecular classification of cutaneous malignant melanoma by gene expression profiling. Nature 2000
- Bittner, Meltzer, et al.
(Show Context)
Citation Context ...a study concerned with the feasibility of producing Markovian networks whose stationary distributions closely reflect the data [10]. The chosen genes arose from data in a study of metastatic melanoma =-=[2]-=-. In this study, the abundance of messenger RNA for the gene WNT5A was found to be highly discriminating between cells with properties typically associated with high metastatic competence versus those... |

122 | Genomic cis-regulatory logic: experimental and computational analysis of a sea urchin gene
- Yuh, Bolouri, et al.
- 1998
(Show Context)
Citation Context ...r being an attractor can be precisely determined for a given model. Regarding the use of the Boolean framework, we note that the logical character of gene regulation has been recognized for some time =-=[18]-=-, [4], [8] and the dynamical behavior of Boolean networks can be used to model many biologically phenomena, such as cellular state dynamics possessing switchlike behavior, stability, and hysteresis [7... |

120 |
A genomic regulatory network for development
- Davidson, Rast, et al.
(Show Context)
Citation Context ...g an attractor can be precisely determined for a given model. Regarding the use of the Boolean framework, we note that the logical character of gene regulation has been recognized for some time [18], =-=[4]-=-, [8] and the dynamical behavior of Boolean networks can be used to model many biologically phenomena, such as cellular state dynamics possessing switchlike behavior, stability, and hysteresis [7]. Fr... |

91 |
Inferring qualitative relations in genetic networks and metabolic arrays
- Akutsu
- 2000
(Show Context)
Citation Context ...BNs offer the potential to design treatment strategies based on the application of external control variables to drive network dynamics [16], [12]. A key issue is network design (inference) from data =-=[1]-=-, [11]. Network connectivity and transition rules must be inferred, with perhaps the imposition of biological constraints [19]. When building function-based gene networks from expression data, the fun... |

75 |
expression profiling, genetic networks, and cellular states: An integrating concept for tumorigenesis and drug discovery
- Huang
- 1999
(Show Context)
Citation Context ...8], [4], [8] and the dynamical behavior of Boolean networks can be used to model many biologically phenomena, such as cellular state dynamics possessing switchlike behavior, stability, and hysteresis =-=[7]-=-. From almost the inception of microarray analysis, logical relations among genes have been constructed from the data [5], [9] and recently the manifestation of logical relations in the continuous dat... |

57 | Gene perturbation and intervention in probabilistic Boolean networks
- Shmulevich, Dougherty, et al.
- 2002
(Show Context)
Citation Context ... 2006 IEEE chains, thereby facilitating steady-state analysis. PBNs offer the potential to design treatment strategies based on the application of external control variables to drive network dynamics =-=[16]-=-, [12]. A key issue is network design (inference) from data [1], [11]. Network connectivity and transition rules must be inferred, with perhaps the imposition of biological constraints [19]. When buil... |

52 | Binary analysis and optimization-based normalization of gene expression data
- Shmulevich, Zhang
- 2002
(Show Context)
Citation Context ... such as tumors and normal tissue, as well as different stages of tumor development, depends on the bimodality of their behavior. The potential for binary discrimination has been shown for clustering =-=[17]-=- and classification [21]. The former has a good discussion of binarization. When using microarray data, which integrates expression over a collection of cells, it should be recognized that we are mode... |

48 |
The Foundations of Science
- Poincaré
- 1913
(Show Context)
Citation Context ...hat a complete biochemical description of cellular activity would likely produce the corollary description of the information processing system does not denigrate the reality of the latter.” Poincaré =-=[14]-=- has remarked: “What we call ‘objective reality’ is, strictly speaking, that which is common to several thinking beings and might be common to all; this common part, we shall see, can only be the harm... |

27 |
Wnt5a signaling directly affects cell motility and invasion of metastatic melanoma
- Weeraratna, Jiang, et al.
(Show Context)
Citation Context ...riminating between cells with properties typically associated with high metastatic competence versus those with low metastatic competence. These findings were validated and expanded in a second study =-=[20]-=-. In this study, experimentally increasing the levels of the Wnt5a protein secreted by a melanoma cell line via genetic engineering methods directly altered the metastatic competence of that cellsDOUG... |

26 | Intervention in context-sensitive probabilistic Boolean networks
- Pal
- 2005
(Show Context)
Citation Context ...IEEE chains, thereby facilitating steady-state analysis. PBNs offer the potential to design treatment strategies based on the application of external control variables to drive network dynamics [16], =-=[12]-=-. A key issue is network design (inference) from data [1], [11]. Network connectivity and transition rules must be inferred, with perhaps the imposition of biological constraints [19]. When building f... |

22 | On learning gene regulatory networks under the Boolean network
- Lähdesmäki, Shmulevich, et al.
- 2003
(Show Context)
Citation Context ...ffer the potential to design treatment strategies based on the application of external control variables to drive network dynamics [16], [12]. A key issue is network design (inference) from data [1], =-=[11]-=-. Network connectivity and transition rules must be inferred, with perhaps the imposition of biological constraints [19]. When building function-based gene networks from expression data, the functions... |

20 |
Steady-state probabilities for attractors in probabilistic Boolean networks
- Brun
- 2005
(Show Context)
Citation Context ...he probability of a PBN being in a particular state can be quantitatively expressed given information on the perturbation and switching probabilities, the attractor structure, and the basin structure =-=[3]-=-. Thus, the probability of an observation vector being an attractor can be precisely determined for a given model. Regarding the use of the Boolean framework, we note that the logical character of gen... |

18 | Can Markov chain models mimic biological regulation
- Kim, Li, et al.
- 2002
(Show Context)
Citation Context ...ctivity approach. The ten genes considered here were first identified in a study concerned with the feasibility of producing Markovian networks whose stationary distributions closely reflect the data =-=[10]-=-. The chosen genes arose from data in a study of metastatic melanoma [2]. In this study, the abundance of messenger RNA for the gene WNT5A was found to be highly discriminating between cells with prop... |

16 |
Epistemology of computational biology: mathematical models and experimental prediction as the basis of their validity
- Dougherty, Braga-Neto
- 2006
(Show Context)
Citation Context ... 9, SEPTEMBER 2006 We close with the epistemological question: How is the obtained PBN, or any such high-level system model, related to a real biological system? We comment only briefly, referring to =-=[6]-=- for a more comprehensive account. When considering models of genomic regulation, one might say that there is a decision-making layer and a physical (chemical) layer. PBNs, and other such functional n... |

16 | A Bayesian connectivity-based approach to constructing probabilistic gene regulatory networks. Bioinformatics
- Zhou, Wang, et al.
(Show Context)
Citation Context ...oposed for PBNs is based on the coefficient of determination [15]. Other methods have been proposed, including optimizing the connectivity of the network according to the data in a Bayesian framework =-=[22]-=-. Except in rare circumstances, the optimal function for a gene will not be a perfect predictor because there will be inconsistencies in the data. This means that a specific vector of values for a set... |

14 |
JM: A General Framework for the Analysis of Multivariate Gene Interaction via Expression Arrays. Biomedical Optic 2000, 5(4):411-424. et al
- Kim, ER, et al.
(Show Context)
Citation Context ... state dynamics possessing switchlike behavior, stability, and hysteresis [7]. From almost the inception of microarray analysis, logical relations among genes have been constructed from the data [5], =-=[9]-=- and recently the manifestation of logical relations in the continuous data has been analyzed [13]. Besides the fact that we are concerned in many of our applications with ON–OFF-type behavior, an imp... |

11 | M.J.T.: ‘Multi-criterion optimization for genetic network modeling
- SOMEREN, WESSELS, et al.
(Show Context)
Citation Context ...k dynamics [16], [12]. A key issue is network design (inference) from data [1], [11]. Network connectivity and transition rules must be inferred, with perhaps the imposition of biological constraints =-=[19]-=-. When building function-based gene networks from expression data, the functions are typically derived via some optimization-based criterion. This requires determining, for each gene , the genes that ... |

6 |
Binarization of Microarray Data Based on a Mixture Model
- Zhou, Wang, et al.
(Show Context)
Citation Context ...al tissue, as well as different stages of tumor development, depends on the bimodality of their behavior. The potential for binary discrimination has been shown for clustering [17] and classification =-=[21]-=-. The former has a good discussion of binarization. When using microarray data, which integrates expression over a collection of cells, it should be recognized that we are modeling global behavior, no... |

4 |
Proceeding of Nonlinear filters in genomic control
- Dougherty, Bittner, et al.
- 1999
(Show Context)
Citation Context ...lular state dynamics possessing switchlike behavior, stability, and hysteresis [7]. From almost the inception of microarray analysis, logical relations among genes have been constructed from the data =-=[5]-=-, [9] and recently the manifestation of logical relations in the continuous data has been analyzed [13]. Besides the fact that we are concerned in many of our applications with ON–OFF-type behavior, a... |

2 |
Boolean relationships among genes responsive to ionizing radiation
- Pal, FornaceJr, et al.
- 2005
(Show Context)
Citation Context ...ception of microarray analysis, logical relations among genes have been constructed from the data [5], [9] and recently the manifestation of logical relations in the continuous data has been analyzed =-=[13]-=-. Besides the fact that we are concerned in many of our applications with ON–OFF-type behavior, an important practical reason for working in the binary setting, or at least in the context of a very co... |