A special attention and a particular design should be considered when one wants to collect expression data not to answer a specific scientific question, but with the idea of constructing a library of expression values under a standard set of circumstances, that subsequent researchers may consult. One of the most important issues in the construction of such library is the identification of a standard set of relevant conditions under which measure expression levels and the definition of standard experimental procedures. Once such standardized expression levels are constructed, one could use statistically based scoring systems (analogous to BLAST) to identify the proximal genes to any novel entry.
References
http://www.ncbi.nlm.nih.gov/BLAST/