Current RNA chip technology, although quite advanced, is usually limited to known transcripts. As rare transcripts (N=1 to 5) usually cannot be quantified, current chip technology is probably useless for building realistic virtual cells. Maybe there are there are other options? SAGE – serial analysis of gene expression – has been also around for some time but never reached a larger audience as the 18 bp tag length did not make it very specific and the read-out of sequencing is not very cost efficient.
This may change now with SuperSAGE having 26 bp tags: cDNA is cut with the tagging enzyme Bsmfl and ligated with a neighboring tag to a ditag; ditags are being sequenced with some of the new technologies like pyro/FLX, bridge pcr/Solexa or solid/ABI. The power certainly comes with the accuracy but also the inclusion of non coding RNAs.