Email updates

Keep up to date with the latest news and content from BMC Bioinformatics and BioMed Central.

Open Access Highly Accessed Research article

Trees on networks: resolving statistical patterns of phylogenetic similarities among interacting proteins

William P Kelly12 and Michael PH Stumpf123*

Author Affiliations

1 Centre for Bioinformatics, Imperial College, London, UK

2 Centre for Integrative Systems Biology at Imperial College (CISBIC), London, UK

3 Institute of Mathematical Sciences, Imperial College, London, UK

For all author emails, please log on.

BMC Bioinformatics 2010, 11:470  doi:10.1186/1471-2105-11-470

Published: 20 September 2010



Phylogenies capture the evolutionary ancestry linking extant species. Correlations and similarities among a set of species are mediated by and need to be understood in terms of the phylogenic tree. In a similar way it has been argued that biological networks also induce correlations among sets of interacting genes or their protein products.


We develop suitable statistical resampling schemes that can incorporate these two potential sources of correlation into a single inferential framework. To illustrate our approach we apply it to protein interaction data in yeast and investigate whether the phylogenetic trees of interacting proteins in a panel of yeast species are more similar than would be expected by chance.


While we find only negligible evidence for such increased levels of similarities, our statistical approach allows us to resolve the previously reported contradictory results on the levels of co-evolution induced by protein-protein interactions. We conclude with a discussion as to how we may employ the statistical framework developed here in further functional and evolutionary analyses of biological networks and systems.