Figure 2.

Experimental frequency of observation of SAGE-tags in yeast. The odds of observing a particular tag from experiment were derived from existing SAGE data for yeast upon eight different culture conditions (Table 2). From left to right in the graphs, white bars correspond to the data Var-1, Var-2 and Var-3 obtained by Varela and collaborators [17]; light gray bars correspond to the data Vel-1, Vel-2 and Vel-3 obtained by Velculescu and collaborators [15]; and dark gray bars correspond to the data Kal-1 and Kal-2 obtained by Kal and collaborators [16]. Average values from the eight data points are provided and also displayed with a black dot. (A) The odds of observing a 'non-poly(A) next' platinum tag upon its position within the transcript were derived from existing experimental SAGE data. The selected experimental tags consisted of those that were unique in the genome, mapped into known transcripts with a predicted 3'-UTR and were not categorized as 'poly(A)-next'. Platinum tags mapping into transcripts without a predicted 3'-UTR were not included to avoid errors of position estimation. (B) The odds of observing a 'poly(A) next' platinum tag was derived from existing experimental SAGE data (left panel). The selected experimental tags consisted of those that were unique in the genome, mapped into known transcripts and were categorized as 'poly(A)-next'. The odds of observing a tag mapping to intergenic positions in the genome was derived from existing SAGE data (right panel). The selected experimental tags consisted of those belonging to the classes copper and iron (Table 1).