Latino Studies at New York University

Karen Lam

Doctoral Candidate
Computational Biology Program
Department of Biology
Courant Institute of Mathematical Sciences
New York University

April 16, 2013

Using global network alignment for multi-species network alignment

There is great diversity in heart morphology within the animal kingdom. However, studies have shown that there are core gene regulatory networks conserved across vast phylogenetic distances.  Ciona intestinalis, an invertebrate chordate, is a good model organism for network inference of heart development because of its cellular and genetic simplicity, but there is a limited amount of data available. We would like to take advantage of the conservation of networks by using readily available mouse data to constrain the Ciona network, and vice versa. The Inferelator algorithm allows us to incorporate prior information in order to improve network inference. We propose using a global network alignment algorithm, along with sequence orthology, to identify the conserved subnetwork for these two species. We can then use this subnetwork to identify cross-species priors on regulatory interactions between genes, so that the network inference algorithm can be repeated with the additional constraints in order to obtain more accurate networks. This technique has the potential to both improve network inference accuracy for both species, and to obtain information about shared regulatory structure.