Figure 7From: Evaluation and integration of functional annotation pipelines for newly sequenced organisms: the potato genome as a test caseRank-based consolidation of the different figures of merit. A non-parametric rank-based consolidation of the different scores of the annotation methods was used for a unified comparison. First, six rankings were calculated: by gene coverage, by NGPG, by the two F-measures of the gold standard validation, and by the two gene co-expression validations scores (i.e., the number and the percent of significant GO terms). To merge the different rankings we used the average rank. The results show that both for PGSC (panel A) and for ITAG (panel B), Ensemble-2 has the best average rank.Back to article page