## Advanced Technologies (Cambridge) Limited

### BibTeX

@MISC{Thaliana_advancedtechnologies,

author = {Of Arabidopsis Thaliana and Rónán Daly and Kieron D. Edwards and John S. O’neill and Stuart Aitken and Andrew J. Millar and Mark Girolami},

title = {Advanced Technologies (Cambridge) Limited},

year = {}

}

### OpenURL

### Abstract

Abstract. Modelling gene regulatory networks in organisms is an important task that has recently become possible due to large scale assays using technologies such as microarrays. In this paper, the circadian clock of Arabidopsis thaliana is modelled by fitting dynamic Bayesian networks to luminescence data gathered from experiments. This work differs from previous modelling attempts by using higher-order dynamic Bayesian networks to explicitly model the time lag between the various genes being expressed. In order to achieve this goal, new techniques in preprocessing the data and in evaluating a learned model are proposed. It is shown that it is possible, to some extent, to model these time delays using a higher-order dynamic Bayesian network.

### Citations

120 | Sensitivity and specificity of inferring genetic regulatory interactions from microarray experiments with dynamic Bayesian networks
- Husmeier
(Show Context)
Citation Context ...onsist of a prior network, used to model initial conditions, and a transition network, used to model the effect of interactions over time. Whilst DBNs have been used to model gene regulatory networks =-=[3,4,5,6,7]-=-, the method presented here differs in how the data are treated and in how it can infer interactions at multiple time periods. To test the methods proposed, a set of gene expression data obtained from... |

49 |
Inferring gene networks from time series microarray data using dynamic Bayesian networks
- Kim, Imoto, et al.
- 2003
(Show Context)
Citation Context ...onsist of a prior network, used to model initial conditions, and a transition network, used to model the effect of interactions over time. Whilst DBNs have been used to model gene regulatory networks =-=[3,4,5,6,7]-=-, the method presented here differs in how the data are treated and in how it can infer interactions at multiple time periods. To test the methods proposed, a set of gene expression data obtained from... |

38 |
D , Chickering DM . Learning Bayesian networks: The combination of knowledge and statistical data
- Heckerman
(Show Context)
Citation Context ...er, most scoring functions for use with structure learning are based on nominal data, i.e. data that is of a discrete nature. For the purposes of these experiments, the BDeu scoring function was used =-=[13]-=-. For this to be utilised, the data has to be discretised. Whilst binary discretisation is an easy option, there is a problem with this as shown by Fig. 2a. It can be seen that as the expression level... |

37 |
Current approaches to gene regulatory network modelling
- Schlitt, Brazma
- 2007
(Show Context)
Citation Context ...refore, methods involving graphical representations have become more widely used. These include Boolean networks, networks similar to neural networks, stochastic process calculi and Bayesian networks =-=[10]-=-. 2.1 Data from Arabidopsis Thaliana Experiments The experimental study in this chapter will involve luminescence data obtained from experiments on Arabidopsis thaliana. The luminescence is an indirec... |

36 | Experimental validation of a predicted feedback loop in the multi-oscillator clock of Arabidopsis thaliana
- Locke, Kozma-Bognar, et al.
- 2006
(Show Context)
Citation Context ...a obtained from experiments observing particular genes of the plant Arabidopsis thaliana will be used. These genes are known to behave in a clock-like fashion that regulates the function of the plant =-=[8,9]-=-. The data will be preprocessed and experiments will be conducted to learn the structure of DBNs using these preprocessed data. Using a new technique, these networks will be compared to a standard net... |

31 |
Conzen SD. A new dynamic Bayesian network (DBN) approach for identifying gene regulatory networks from time course microarray data
- Zou
(Show Context)
Citation Context ...onsist of a prior network, used to model initial conditions, and a transition network, used to model the effect of interactions over time. Whilst DBNs have been used to model gene regulatory networks =-=[3,4,5,6,7]-=-, the method presented here differs in how the data are treated and in how it can infer interactions at multiple time periods. To test the methods proposed, a set of gene expression data obtained from... |

25 | On the dirichlet prior and bayesian regularization
- Steck, Jaakkola
- 2002
(Show Context)
Citation Context ...hould perform well in most cases [1]. It has been known for some time that the value of the equivalent sample size parameter N ′ for the BDeu scoring function has a large effect on learning structure =-=[15,16,17]-=-. In a sense, the value of N ′ can be seen as a regularisation parameter - the larger the value of N ′ , the more edges are supported in the learned graph [18]. This can be contrasted to scores like t... |

24 |
Reconstructing gene-regulatory networks from time series, knock-out data, and prior knowledge
- Geier, Timmer, et al.
(Show Context)
Citation Context |

12 | Learning Bayesian network equivalence classes with ant colony optimization
- Daly, Shen
(Show Context)
Citation Context ...ular, the time lag between the expression of a gene and its effect on the expression of another gene. This modelling will proceed by using an algorithm based on ant colony optimisation known as ACO-E =-=[1]-=- to build dynamic Bayesian networks (DBNs). DBNs are an extension of Bayesian networks, graphical models that can be useful in modelling probabilistic relations between variables [2]. They consist of ... |

9 |
high-temperature response of the Arabidopsis circadian clock
- Edwards, Anderson, et al.
- 2006
(Show Context)
Citation Context ...a obtained from experiments observing particular genes of the plant Arabidopsis thaliana will be used. These genes are known to behave in a clock-like fashion that regulates the function of the plant =-=[8,9]-=-. The data will be preprocessed and experiments will be conducted to learn the structure of DBNs using these preprocessed data. Using a new technique, these networks will be compared to a standard net... |

7 | On sensitivity of the MAP Bayesian network structure to the equivalent sample size parameter
- Silander, Kontkanen, et al.
- 2007
(Show Context)
Citation Context ...hould perform well in most cases [1]. It has been known for some time that the value of the equivalent sample size parameter N ′ for the BDeu scoring function has a large effect on learning structure =-=[15,16,17]-=-. In a sense, the value of N ′ can be seen as a regularisation parameter - the larger the value of N ′ , the more edges are supported in the learned graph [18]. This can be contrasted to scores like t... |

7 | A bayesian network scoring metric that is based on globally uniform parameter priors
- Kayaalp, Cooper
- 2002
(Show Context)
Citation Context ...hould perform well in most cases [1]. It has been known for some time that the value of the equivalent sample size parameter N ′ for the BDeu scoring function has a large effect on learning structure =-=[15,16,17]-=-. In a sense, the value of N ′ can be seen as a regularisation parameter - the larger the value of N ′ , the more edges are supported in the learned graph [18]. This can be contrasted to scores like t... |

5 |
d’Alché Buc, Gene networks inference using dynamic bayesian networks, Bioinformatics 19
- Perrin, Ralaivola, et al.
- 2003
(Show Context)
Citation Context |

4 |
Aitken S: Learning Bayesian networks: approaches and issues
- Daly, Shen
(Show Context)
Citation Context ...on known as ACO-E [1] to build dynamic Bayesian networks (DBNs). DBNs are an extension of Bayesian networks, graphical models that can be useful in modelling probabilistic relations between variables =-=[2]-=-. They consist of a prior network, used to model initial conditions, and a transition network, used to model the effect of interactions over time. Whilst DBNs have been used to model gene regulatory n... |

4 | Learning the bayesian network structure: Dirichlet prior vs. data
- Steck
- 2008
(Show Context)
Citation Context ...large effect on learning structure [15,16,17]. In a sense, the value of N ′ can be seen as a regularisation parameter - the larger the value of N ′ , the more edges are supported in the learned graph =-=[18]-=-. This can be contrasted to scores like the BIC and AIC, where regularisation is implicit in the function and cannot be adjusted. Therefore, inUsing Higher-Order Dynamic Bayesian Networks to Model Pe... |

3 |
Large-scale regulatory network analysis from microarray data: modified Bayesian network learning and association rule mining, Decision Support Systems
- Huang, Jiexun, et al.
(Show Context)
Citation Context ...tic model. This enables large scale, non-linear interactions between many genes to be represented and simulated. E.g., recent studies have looked at learning Bayesian networks with thousands of genes =-=[12]-=-. The large amounts of data present in microarray studies can be used to learn both the structure and parameters of the network – missing values and noise can also be taken care of. Also, Bayesian net... |

2 |
D.: Modeling multiple time units delayed gene regulatory network using dynamic Bayesian network
- Xing, Wu
- 2006
(Show Context)
Citation Context ...ossible dynamic Bayesian network was set to 9, i.e. the time slice at t and 8 slices back. Other authors show using DBNs with multiple layers to model gene regulatory networks with multiple time lags =-=[14]-=-. However, their methods consider arcs between arbitrary layers that are not at time t. This method is incorrect, as it means that arcs could be added between nodes without taking into account the oth... |

1 |
Plant circadian rhythms. The Plant Cell 18
- McClung
- 2006
(Show Context)
Citation Context ...30 AT0031b AT0032 AT0033 AT0047 Ratio 6:18 9:15 12:12 15:9 18:6 3:21 the ‘plant circadian rhythms’ of this organism, i.e. the oscillating behaviour of plants as they respond to the change in sunlight =-=[11]-=-. Briefly, the expression levels of the genes in plants tend to synchronise with the rising and setting of the sun, oscillating between high and low levels. However, when the light source is removed, ... |