Skip to main content
Figure 3 | BMC Bioinformatics

Figure 3

From: Combining gene prediction methods to improve metagenomic gene annotation

Figure 3

Positive Predictive Value of of the Orphelia, MetaGeneAnnotator, and GeneMark. The specificity computation (TP/(TP+FP) used in Hoff et al., or positive predictive value, of the three programs, Orphelia, MGA and GeneMark. This metric measures the programs' precision to correctly predict genes, as opposed to the traditional specificity value which measures a test's ability to correctly identify non-coding regions. The values here reflect similar "specificity" values found in the literature [23], [22]. The PPV metric does not fluctuate as much as the specificity because it takes into account the bias of the coding-to-non-coding ratio (75%/25%) of our simulated dataset. On the other hand, the specificity metric assesses algorithms' true ability to detect non-coding regions. In this case, GeneMark has the best PPV for short reads, while GeneMark and Orphelia are comparable for longer reads.

Back to article page