Table 1 Average test scores (± standard deviation) obtained by the different models in the random hold out use case (UC1). The suffix +emb indicates that protein sequence embeddings were concatenated to the I/O features. Best results are boldfaced
From: Sensitivity analysis on protein-protein interaction networks through deep graph networks
Model | MCC | F1 | ACC | AUROC |
|---|---|---|---|---|
Null | \(.000{\scriptstyle \pm .000}\) | \(.000{\scriptstyle \pm .000}\) | \(.676{\scriptstyle \pm .000}\) | \(.500{\scriptstyle \pm .000}\) |
DeepSets | \(.062\scriptstyle \pm .043\) | \(.284\scriptstyle \pm .140\) | \(.574\scriptstyle \pm .167\) | \(.650\scriptstyle \pm .004\) |
DeepSets+emb | \(.472\scriptstyle \pm .006\) | \(.647\scriptstyle \pm .004\) | \(.764\scriptstyle \pm .004\) | \(.817\scriptstyle \pm .002\) |
DGN | \(.625\scriptstyle \pm .009\) | \(.750\scriptstyle \pm .006\) | \(.830\scriptstyle \pm .009\) | \(.907\scriptstyle \pm .004\) |
DGN+emb | \(\mathbf {.767\scriptstyle \pm .004}\) | \(\mathbf {.843\scriptstyle \pm .002}\) | \(\mathbf {.897\scriptstyle \pm .003}\) | \(\mathbf {.957\scriptstyle \pm .002}\) |