Skip to main content

Table 9 Comparison of the average AUC, precision at 20% recall (P20R) and precision at 40% recall between multi-species and single-species approaches with 301 species of bacteria

From: Think globally and solve locally: secondary memory-based network learning for automated multi-species function prediction

Multi-species approach

Algorithm

AUC

P20R

P40R

RW - 1 step

0.8744

0.2264

0.1673

RW - 2 steps

0.8590

0.1318

0.0893

RW - 3 steps

0.8419

0.1064

0.0713

Single-species approach

Algorithm

AUC

P20R

P40R

RW - 1 step

0.8263

0.1801

0.1176

RW - 2 steps

0.8146

0.1059

0.0647

RW - 3 steps

0.8179

0.1009

0.0563