2.4. Model validation

To validate the performance of our algorithm, each individual PPI from the positive dataset was taken out of the predictive models and used as query input (leave-one-out cross-validation). All the PPIs recovered in the combined interactome generated were considered True Positives (TP), while the rest were considered False Negatives (FN). Subsequently, we passed each of the non-interacting pairs of proteins as input to our predictive algorithm. In this case, any non-interacting pair recovered in the combined interactome was treated as a False Positive (FP), whereas the remaining ones were treated as True Positives (TP). Evaluation metrics were calculated as: