S18-02 Inferring molecular networks

MECHANISMS OF DEVELOPMENT

1 2 6 ( 2 0 0 9 ) S 1 8 –S 1 9

available at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/modo

Symposium – Modelling and networks S18-01

computationally predict various molecular networks during other

Scaling of morphogen gradients during embryonic development

biological or pathological processes.

Naama Barkai Weizamnn Institute of Science, Rehovot, Israel The ability to maintain a proportionate body plan in individuals of different sizes is essential for proper development. Despite wide interest, little is known about the mechanisms underlying scaling with size. I will discuss our efforts to reveal the molecular

doi:10.1016/j.mod.2009.06.991

S18-03 Shaping the early amniote embryo: An object-oriented model based on defined cell behaviours Octavian Voiculescu1, Claudio Stern1, Lawrence Bodenstein2

basis for scaling in different systems. A general ‘‘Expansion–

1

University College London, London, United Kingdom

Repression’’ feedback topology will be presented, which ensures

2

Olana Technologies, New York, United States

scaling for a wide range of parameters and sizes through an effective implementation of an integral-feedback controller. Integral-

We have previously investigated the individual cell behaviours

feedback presents a key concept in engineering, and is likely to

of the epiblast of chick embryos at pre- and gastrula-stages, and

be instrumental also for maintaining biological homeostasis.

shown that (1) cell intercalation occurs well before gastrulation

doi:10.1016/j.mod.2009.06.990

in the prospective domain of the primitive streak (Voiculescu etal., 2007), and (2) cells in the whole epiblast ingress at a very low rate but show a very strong cooperativity in the primitive streak (hence, ingression is mainly achieved there – see abstract by Voiculescu,

S18-02

Lau & Stern). Based exclusively on these experimental data, we

Inferring molecular networks

describe an object-oriented model of the early chick embryo. First,

Jing-Dong Jackie Han

we show that these components are sufficient to account for the morphogenetic movements in the epiblast at pre-primitive streak

Chinese Academy of Sciences, Beijing, China

stages (‘Polonaise movements’, Gra¨per, 1929; Spratt, 1946) and during gastrulation (Wetzel, 1929) in the model. Second, we confront

With the increasing quantities of high throughput data, it is

the experimental observations with computer simulations when

possible to infer biological networks from these measurements.

one of these components is abrogated. Finally, we discuss the pre-

Undirected networks can be inferred by associations or correla-

dictive capacity of the model to explain embryological data for

tions between variables (be these genes or other biological enti-

which a mechanistic explanation has been elusive.

ties), whereas directed networks capture the directionality of the relationships between variables and can be inferred by the conditional probability distributions of variables. These relationships may correspond to causal relationships between the variables, thus a model A B may indicate that the occurrence of node A promotes or prohibits that of node B. For example, recent high throughput technologies, such as ‘ChIP-chip’ and ‘ChIP-seq’,

References:

Voiculescu O, Bertocchini F, Wolpert L, Keller RE, Stern CD. Nature 2007;449:1049. Gra¨per L. Arch. EntwMech. Org. 1929;116:382. Spratt NT. J. Exp. Zool. 1946;103:259. Wetzel R. Arch. EntwMech. Org. 1929;119:188.

have generated high resolution maps for many histone modifications on the human genome. Different modifications may com-

doi:10.1016/j.mod.2009.06.992

bine to form complex ‘histone codes’. To infer these codes, we used the Bayesian network framework to find causal and combinatorial relationships among histone modifications and gene expression. Our unbiased network model not only confirmed already known relationships, such as those of H3K27me3 to gene

S18-04 Synchronization of oscillating cells during embryogenesis Andrew Oates

silencing, H3K4me3 to gene activation, and the effect of bivalent

Max Planck Institute of Molecular Cell Biology and Genetics, Dresden,

modification of both H3K4me3 and H3K27me3, but also identified

Germany

many other relationships that may predict new epigenetic interactions important in epigenetic gene regulation. We also devel-

The sequential formation of somites along the elongating

oped and applied network structure inference methods to

zebrafish body axis is a model system for understanding the