Figure 3.
Precision and recall for different reliability indices (RIs). We defined a reliability index (RI) measuring the strength of a prediction (Eq. 5), i.e., for the embedding proximity. Precision (Eq. 1) and recall (Eq. 2) were almost constant for lower RIs (up to ~ 0.3) for all three ontologies (BPO, MFO, CCO). For higher RIs, precision increased while recall dropped. However, due to the low number of terms predicted at very high RIs (> 0.8), precision fluctuated and was not fully correlated with RI. Panel (a) shows precision and recall for BPO, panel (b) for MFO, and panel (c) for CCO. Dashed vertical lines marked the thresholds used by the CAFA3 tool to compute Fmax: 0.35 for BPO, 0.28 for MFO, and 0.29 for CCO. At least for BPO and MFO higher RI values correlated with higher precision, i.e., users could use the RI to estimate how good the predictions are likely to be for their query, or to simply scan only those predictions more likely to be correct (e.g. RI > 0.8).