Most gene functions of biochemical pathways were still unexplored, especially interactions of constituent genes. We attempted to uncover interaction network of biochemical pathways via a survey of co-expression clusters, which we have constructed from the NCBI GEO database, and then to define key genes of networks with expression correlations between members. Top 20 pathways with high numbers of individual genes were retrieved from 178 pathways. One pathway, ‘removal of superoxide radicals’ was excluded for further study, evidencing somewhat low degree (16%, 13 out of 79 genes) of mapped probes. We employed expression correlations of random pairs of 1,000 randomly selected genes for determining a cut off r-value for gene networks. Numbers of interactions with a significant expression correlation values between members might evidence that “hub genes” play key roles among a given pathway genes. For example most interactive pathway, ‘tRNA charging pathway’, that is composed of 60 probes corresponding to genes showed 264 positive significant interactions between members of 47 genes while 5 negative interactions between members of 7 genes., evidencing ‘Os10g26050’ (methionyl-tRNA synthetase) gene with highest interactions is suggestive of a hub gene. These findings might provide some clues on evolutionary fate of co-expression genes including each of biochemical pathways, e.g. convergent evolution

• Won Cheol Yim(Dept. of Plant Biotechnology, Dongguk Univ.)
• Cheol Seong Jang(Plant Genomics Lab., Div. of Bio-resources Technology, Kangwon Natl. Univ.) Corresponding Author
• Byung-Moo Lee(Dept. of Plant Biotechnology, Dongguk Univ.) Corresponding Author