We have seen how CAGE and tiling microarray analysis support the identification of several unexpected transcripts and, consequently, CAGE data can identify the TSS and promoter for each of them.

In some case, this gave extremely puzzling results, since the transciption unit identified do not resemble any of the previously known TU.

Could you suggest methods to test in the laboratory if the promoters identified at the 5' of these unusual TU are real promoters ? Make an effort to resume the method described in Kim et al. (in the lecture) and suggest a possible additional method.