Does anyone know whether the interactions are conserved over long-evolutionary time periods (> 300 mya). For example, if two genes are interacting in one species, will the orthologs of these two genes be still interacting in another species? Any good reference on this.
I guess you are refering to the concept of "interologs". There are some papers out there (e.g. http://www.ncbi.nlm.nih.gov/pubmed/17535438) showing that many interactions (not all) are conserved along evolution.
There are two related concepts that you may want to check out: