A Transformer-Based Ensemble Framework for the Prediction of Protein–Protein Interaction Sites

Abstract

The identification of protein–protein interaction (PPI) sites is essential in the research of protein function and the discovery of new drugs. So far, a variety of computational tools based on machine learning have been developed to accelerate the identification of PPI sites. However, existing methods suffer from the low predictive accuracy or the limited scope of application. Specifically, some methods learned only global or local sequential features, leading to low predictive accuracy, while others achieved improved performance by extracting residue interactions from structures but were limited in their application scope for the serious dependence on precise structure information. There is an urgent need to develop a method that integrates comprehensive information to realize proteome-wide accurate profiling of PPI sites. Herein, a novel ensemble framework for PPI sites prediction, EnsemPPIS, was therefore proposed based on transformer and gated convolutional networks. EnsemPPIS can effectively capture not only global and local patterns but also residue interactions. Specifically, EnsemPPIS was unique in (a) extracting residue interactions from protein sequences with transformer and (b) further integrating global and local sequential features with the ensemble learning strategy. Compared with various existing methods, EnsemPPIS exhibited either superior performance or broader applicability on multiple PPI sites prediction tasks. Moreover, pattern analysis based on the interpretability of EnsemPPIS demonstrated that EnsemPPIS was fully capable of learning residue interactions within the local structure of PPI sites using only sequence information. The web server of EnsemPPIS is freely available at http://idrblab.org/ensemppis .

Large Language Model-Based Natural Language Encoding Could Be All You Need for Drug Biomedical Association Prediction

“…The test PRC curves and ROC curves among 10-folds are shown in Figure a,b. Furthermore, the results of our strategy were compared with the other typical tools using nonparametric one-side Mann–Whitney U tests. , The alternative hypothesis was defined as our performance values being not higher than the others. As shown in Figure c–e, our strategy was especially advantageous at AUPR and always ranked top 1, and consistently ranked top 3 at AUROC and ACC (the results of the contrastive tools were from the reports in DREAMwalk under identical experimental settings).…”

MultiClassMetabo: A Superior Classification Model Constructed Using Metabolic Markers in Multiclass Metabolomics

“…Clustering analysis was adopted to indicate the separation degree of samples in multiple classes using metabolic markers for multiclass metabolomic data. , For a clustering outcome, a method of identifying metabolic markers was regarded as superior when an obvious separation was observed for different sample classes. A well-established metric ( purity ) in the clustering analysis was used as a representative measure to assess the quality of clustering .…”

MOINER: A Novel Multiomics Early Integration Framework for Biomedical Classification and Biomarker Discovery

“…35 This limitation hinders the application of such data in advanced deep learning models, including 2D-CNN and Vision Transformer. 36,37 To address these challenges, we propose MOINER, a novel multi-omics early integration framework based on information enhancement and image representation learning strategies. Specifically, all feature variables within the raw high-dimensional multiomics data are designated as a global feature set (GFS), while the feature subsets resulting from feature selection are designated as a local feature set (LFS).…”

“…(2) Another challenge lies in the fact that sequential high-dimensional multiomics vectors can hardly reflect the intrinsic correlations of omics-features from the representational level . This limitation hinders the application of such data in advanced deep learning models, including 2D-CNN and Vision Transformer. , …”