Are you using the clusters to create a data matrix with which to create a phylogenetic tree? If that's the case then I would recommend collecting the universally present single copy clusters (clusters with only one member i.e putative true orthologs) and creating a multiple sequence alignment with those as an input.

If you already have a tree and you're trying to look at the content of the groups in the context of the phylogeny then there's quite a few programs out there to evaluate ancestral traits over a tree. You could try BayesTraits http://www.evolution.rdg.ac.uk/BayesTraits.html as one possible program or maybe something from the PHYLIP suite.

evolution.genetics.washington.edu/phylip.htm

Hope that helps,

Saif

Thanks for your reply. I have a total of 103 clusters shared by 7 species. I could n't get it completely. Do I have to generate MSA for all of there 103 clusters?

To construct a tree, it's probably better to use to clusters which only have 1 member per cluster shared by all taxa. Those proteins are more likely to be orthologous (i.e. inherited from the last common ancestor of all the taxa under consideration).