InferPPI aims to infer protein-protein interaction based on the parsimony principle of domain-domain interactions. Specifically, inferring domain interactions is a first step and formulated as an integer linear programming with different searching degree of training protein interaction. Then the model is solved by relaxed linear programming. By using the obtained domain interactions, novel protein interactions can be predicted by a formula. In contrast to probabilistic algorithms, InferPPI is an algorithm based on a deterministic model.
:: MORE INFORMATION
Xiang-Sun Zhang, Rui-Sheng Wang, Ling-Yun Wu, Yong Wang, Shihua Zhang, Luonan Chen.
Predicting protein interactions based on parsimony tendency of domain interactions.