Elucidating protein-protein interactions (PPIs) is essential to constructing protein in-teraction networks and facilitating our understanding of the general principles of bio-logical systems. Previous studies have revealed that interacting protein pairs can be predicted by their primary structure. Most of these approaches have achieved satis-factory performance on datasets comprising equal number of interacting and non-interacting protein pairs. However, this ratio is highly unbalan! ced in nature, and these techniques have not been comprehensively evaluated with respect to the effect of the large number of non-interacting pairs in realistic datasets. Moreover, since highly unbalanced distributions usually lead to large datasets, more efficient predic-tors are desired when handling such challenging tasks.
This study presents a method for PPI prediction based only on sequence information, which contributes in three aspects. First, we propose a probability-based mechanism for transforming protein sequences into feature vectors. Second, the proposed predic-tor is designed with an efficient classification algorithm, where the efficiency is es-sential for handling highly unbalanced datasets. Third, the proposed PPI predictor is assessed with several unbalanced datasets with different positive-to-negative ratios (from 1:1 to 1:15). This analysis provides solid evidence that the de! gree of dataset imbalance is important to PPI predictors.
Dealing with data imbalance is a key issue in PPI prediction since there are far fewer interacting protein pairs than non-interacting ones. This article provides a compre-hensive study on this issue and develops a practical tool that achieves both good pre-diction performance and efficiency using only protein sequence information.