Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 7 (1), 8007

SpotOn: High Accuracy Identification of Protein-Protein Interface Hot-Spots


SpotOn: High Accuracy Identification of Protein-Protein Interface Hot-Spots

Irina S Moreira et al. Sci Rep.


We present SpotOn, a web server to identify and classify interfacial residues as Hot-Spots (HS) and Null-Spots (NS). SpotON implements a robust algorithm with a demonstrated accuracy of 0.95 and sensitivity of 0.98 on an independent test set. The predictor was developed using an ensemble machine learning approach with up-sampling of the minor class. It was trained on 53 complexes using various features, based on both protein 3D structure and sequence. The SpotOn web interface is freely available at: .

Conflict of interest statement

The authors declare that they have no competing interests.


Figure 1
Figure 1
Cluster dendrogram of the machine learning algorithms tested in this work. All 5 clusters are separated by a dashed line and are ordered from I to V.
Figure 2
Figure 2
Collage of the results page of the SpotOn webserver: Screenshot of the webGL structure viewer highlighting the hot spot residues in the interface (top); table listing the residues classified as HS (middle) and; sequence viewer highlighting the residues classified as HS and NS in the full sequence of the chains submitted for analysis (bottom).
Figure 3
Figure 3
Workflow of the SpotOn web server pipeline. Each box corresponds to a step in the pipeline and the horizontal bars at the bottom of the image indicate the environment in which this step takes place. At the very beginning, the user is required to upload the PDB file in addition to defining the two monomers of the interface. After the credentials of the user have been checked and the input data validated, the web server will generate the run directory with all the necessary files. In case of validation errors, a helpful message is displayed on screen indicating the exact problem. The master node of the Linux cluster where SpotOn is hosted monitors the directory where the run folders are located and if the global maximum number of concurrent SpotOn jobs or the maximum number of jobs per user have not exceeded the defined limits, the analysis is submitted to the queue. Depending on the load of the system at the time of submission, the analysis might start running immediately or with a small delay. The user is notified as soon as the job starts running. The actual run takes place in one of the working nodes of the cluster and, upon completion, the user is notified via email.

Similar articles

See all similar articles

Cited by 6 PubMed Central articles

See all "Cited by" articles


    1. Petta I, Lievens S, Libert C, Tavernier J, De Bosscher K. Modulation of Protein-Protein Interactions for the Development of Novel Therapeutics. Mol. Ther. 2016;24:707–718. doi: 10.1038/mt.2015.214. - DOI - PMC - PubMed
    1. Clackson T, Wells JA. A hot spot of binding energy in a hormone-receptor interface. Science. 1995;267:383–386. doi: 10.1126/science.7529940. - DOI - PubMed
    1. Moreira IS. The Role of Water Occlusion for the Definition of a Protein Binding Hot-Spot. Curr. Top. Med. Chem. 2015;15:2068–2079. doi: 10.2174/1568026615666150519103733. - DOI - PubMed
    1. Moreira IS, Fernandes PA, Ramos MJ. Hot spots–a review of the protein-protein interface determinant amino-acid residues. Proteins. 2007;68:803–812. doi: 10.1002/prot.21396. - DOI - PubMed
    1. Ramos RM, Moreira IS. Computational Alanine Scanning Mutagenesis-An Improved Methodological Approach for Protein-DNA Complexes. J. Chem. Theory Comput. 2013;9:4243–4256. doi: 10.1021/ct400387r. - DOI - PubMed

Publication types