Home

About Us: Our Halo Research

ISB Halo Group

Educational Outreach

Publications

Raw Data

Genomic Sequence

Regulatory Network

Physical/Functional Network

Related Websites

Picture Gallery

Guest Book

        
Operon/regulon edges: Three edge types aimed at uncovering co-regulation patterns were calculated, two based on chromosomal proximity and one ebased on several microarray  experiments carried out on Halobacterium. We used two simple methods to predict conserved or significant chromosomal proximity. These methods were aimed at uncovering statistically significant chromosomal proximity and not the prediction of operons. One edge type indicates  conserved patterens of proximity. Two proteins were given a chromosomal-proximity link if they were within 300 bp of each other in Halobacterium and had orthologs in at least one other organism within 300 bp. This method has the disadvantage that it cannot give us insight into the function of proteins without orthologs in other systems. For proteins not grouped by the above method we used the method of Moreno-Hagelsieb and Collado-Vides to predict significant proximities on the basis of distance alone, as previously described. Links were generated by these two rough operon prediction methods.

      Co expression over microarray experiments, studies carried out here at the ISB, were used to determine groups of co-regulated genes. Genes were considered coregulated if the Pearson correlation coefficient between the two genes was higher that 0.97 and anti-correlated if below 0.97. These co-regulation edges are then used in combination with chromosomal proximity edges, as described above.  

Operon/regulon edges         Gene-fusion edges         Pylogenetic profile edges              Physical interaction edges        Top

Gene-fusion edges : Enright et al. (1999) have demonstrated that domain fusions are often correlated with functional interactions among the corresponding proteins. Domains within Halobacterium sp. proteins fused in other genomes have been used as a measure to predict
functional interactions between proteins. The co-expression and chromosomal co-localization of the genes will allow identification of true interacting pairs especially in cases where one or more paralogs of the same protein are predicted to interact with a common partner. 

Operon/regulon edges         Gene-fusion edges         Pylogenetic profile edges              Physical interaction edges        Top

Pylogenetic profile edges: Consistent co-localization of pairs of conserved homologs (PCHs) across genomes has been used as the basis for determining functional coupling genes from an evolutionary standpoint (Overbeek et al., 1999).  These functional couplings include proteins that may (i) participate in the same biochemical pathway, (ii) interact with each other, and/or (iii) be co-regulated in response to a common stimulus. To identify these relationships we use the method of Marcotte et al, commonly referred to as the phylogenetic profile method.

Physical interaction edges: We have mapped two-hybrid interaction maps for yeast on to Halobacterium sp. NRC-1 orthologs of yeast and Helicobacter pylori proteins as catalogued in the COG database. Possible protein binding pairs in Halobacterium sp. were inferred in a three step process: first, the COG members of binding pairs in the two hybrid maps were determined; second, all corresponding "COG partners" of the yeast and H. pylori proteins were identified in Halobacterium sp. genome; and finally, the interaction pairs were attributed a confidence level which was computed as a function of the strength of the match to the COG pair and the confidence of the two hybrid edge. In all a total of 1431interactions within the Halobacterium sp. proteome were inferred by this method.

 Operon/regulon edges         Gene-fusion edges         Pylogenetic profile edges              Physical interaction edges        Top