Skip to main content
Figure 4 | BMC Bioinformatics

Figure 4

From: Accurate and unambiguous tag-to-gene mapping in serial analysis of gene expression

Figure 4

Examples of tag confidence assignments by the HGA procedure. Different scenarios of confidence class assignments for virtual genomic tags are shown. In panels A through D, confidence class assignments for aluminum tags are exemplified. In panel E and F, confidence class assignments for tags belonging to different classes are shown. Unique tag sequences always have a high confidence and are only included to provide a genomic context (empty rectangles). Genomic tags sharing the same sequence are illustrated by a dark gray colour. In the case of tags mapping into an annotated transcript, the position within the transcript is also shown. Poly(A) next tags are identified by the presence of a downstream poly(A) region. Before the assignment is performed, the probability of experimental observation for each relevant virtual tag instance is obtained from Table 3, depending on the observed features of each tag. Then, all possible odds ratios using the largest value in the numerator are calculated. In those cases where a particular tag instance exhibits an odd ratio larger than 5.0 when compared against all other genomic instances of the tag sequence, a high confidence class is assigned to that tag. The remaining tag instances are assigned either a low or an undefined confidence class. The undefined confidence class is assigned when the odd ratio between two instances is lower than the threshold of 5.0. A detailed flowchart of the tag confidence assignment process is shown in Figure 3.

Back to article page