Prediction of Hot Spots by Using Structural Neighborhood Properties

PredHS2 About


PredHS2 is a webserver which can effectively predict hot spots on protein-protein interaction interfaces, it is based on our previous method PredHS. Figure 1 shows the flowchart of PredHS2. Firstly, the training dataset is generated by integrating four datasets including ASEdb, SKEMPI, Ab+, and Alexov sDB. And the independent dataset is extracted from the BID database. The residues in the datasets are encoded using a large number of sequence, structure, energy and exposure features, and two categories of structural neighborhood properties (Euclidean and Voronoi). As a result, a total of 200 site features, 200 Euclidean features, and 200 Voronoi features are obtained. Then a two-step feature selection approach is applied to select the optimal feature set. Finally, the prediction classifier is built using Extreme Gradient Boosting based on the optimal feature set.

Figure 1. Flowchart of PredHS2

Current Version

* Current Version Number: 2.0
* Release Date of Current Version: 5/01/2018

Contact Info

E-mail:csu_haowang(at) , leideng(at)


[1].   Deng L, Zhang QC, Chen Z, Meng Y, Guan J and Zhou S. PredHS: a web server for predicting protein–protein interaction hot spots by using structural neighborhood properties. Nucleic Acids Research, 42(W1):W290-W295 (2014) [PubMed]
[2].   Deng L, Guan J, Wei X, Yi Y, Zhang QC and Zhou S. Boosting Prediction Performance of Protein–Protein Interaction Hot Spots by Using Structural Neighborhood Properties. Journal of Computational Biology, 20(11):878–891 (2013) (Presented at RECOMB 2013)[PubMed]
[3].   H Wang, C Liu, L Deng. Enhanced prediction of hot spots at protein-protein interfaces using extreme gradient boosting. Scientific reports, 8 (1):14285 (2018) [PubMed]