Positive unlabeled-learning is a semisupervised machine learning method for scenarios when there is only positive data is labeled, like in fraud detection and disease gene identification. This algorithm is one such approach for disease gene identification.