Improving prediction of heterodimeric protein complexes using combination with pairwise kernel

Abstract

BackgroundSince many proteins become functional only after they interact with their partner proteins and form protein complexes, it is essential to identify the sets of proteins that form complexes. Therefore, several computational methods have been proposed to predict complexes from the topology and structure of experimental protein-protein interaction (PPI) network. These methods work well to predict complexes involving at least three proteins, but generally fail at identifying complexes involving only two different proteins, called heterodimeric complexes or heterodimers. There is however an urgent need for efficient methods to predict heterodimers, since the majority of known protein complexes are precisely heterodimers.ResultsIn this paper, we use three promising kernel functions, Min kernel and two pairwise kernels, which are Metric Learning Pairwise Kernel (MLPK) and Tensor Product Pairwise Kernel (TPPK). We also consider the normalization forms of Min kernel. Then, we combine Min kernel or its normalization form and one of the pairwise kernels by plugging. We applied kernels based on PPI, domain, phylogenetic profile, and subcellular localization properties to predicting heterodimers. Then, we evaluate our method by employing C-Support Vector Classification (C-SVC), carrying out 10-fold cross-validation, and calculating the average F-measures. The results suggest that the combination of normalized-Min-kernel and MLPK leads to the best F-measure and improved the performance of our previous work, which had been the best existing method so far.ConclusionsWe propose new methods to predict heterodimers, using a machine learning-based approach. We train a support vector machine (SVM) to discriminate interacting vs non-interacting protein pairs, based on informations extracted from PPI, domain, phylogenetic profiles and subcellular localization. We evaluate in detail new kernel functions to encode these data, and report prediction performance that outperforms the state-of-the-art.

An optimized deep learning architecture for the diagnosis of COVID-19 disease based on gravitational search optimization

“…The convolutional base includes three significant types of layers are: convolutional layers [31] , activation layers [32] , and pooling layers [33] . These types of layers are used to discover the basic features of input images, which are called feature maps.…”

PC2P: parameter-free network-based prediction of protein complexes

“…1C ). This observation has led to the design of algorithms to identify sparse ( Srihari and Leong, 2012 ; Yong et al , 2012 ) as well as small complexes ( Ruan et al , 2018 ; Yong et al , 2014 ), which have slightly improved the recall of protein complexes.…”

Efficient and accurate identification of protein complexes from protein-protein interaction networks based on the clustering coefficient

“…In addition, comparative analyses have demonstrated that these approaches are not able to predict high-confidence clusters and suffer from small recall [33] . This observation has led to the design of algorithms to identify sparse [34] , [35] and small complexes [36] , [37] , which have slightly improved the recall of protein complexes. Yet, these algorithms depend on multiple parameters, which render it difficult to gauge the performance in absence of optimal parameter values for different combinations of PPI networks and gold standards.…”