Figure 4

Flow chart of StudentC. StudentC first counts how often each GO term appeared in the BLAST hits and performs a cumulative propagation: for each inner node, all counts of its child terms are summed up and added to its own count (depth-first traversal). Dividing each count by that of the root term, we obtain a score between 0 and 1 for each term. In parallel, we calculate a second score for each term by assigning it the maximum percentage of positives of all associated hits (see text). Finally, we multiply the two scores, determine the highest scoring leaf term and output only its propagation to the user.