Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
, 44 (W1), W430-5

RaptorX-Property: A Web Server for Protein Structure Property Prediction


RaptorX-Property: A Web Server for Protein Structure Property Prediction

Sheng Wang et al. Nucleic Acids Res.


RaptorX Property ( is a web server predicting structure property of a protein sequence without using any templates. It outperforms other servers, especially for proteins without close homologs in PDB or with very sparse sequence profile (i.e. carries little evolutionary information). This server employs a powerful in-house deep learning model DeepCNF (Deep Convolutional Neural Fields) to predict secondary structure (SS), solvent accessibility (ACC) and disorder regions (DISO). DeepCNF not only models complex sequence-structure relationship by a deep hierarchical architecture, but also interdependency between adjacent property labels. Our experimental results show that, tested on CASP10, CASP11 and the other benchmarks, this server can obtain ∼84% Q3 accuracy for 3-state SS, ∼72% Q8 accuracy for 8-state SS, ∼66% Q3 accuracy for 3-state solvent accessibility, and ∼0.89 area under the ROC curve (AUC) for disorder prediction.


Figure 1.
Figure 1.
Illustration of DeepCNF. Here, i is the position index and Xi the associated input features, Hk represents the kth hidden layer, and Y is the output label. All the layers from the 1st to the Kth form a DCNN with parameter Wk{k = 1,2,…,K}, which is shown in blue. The Kth layer and the label layer form a CRF (shown in red), in which the parameter U specifies the relationship between the Kth layer and the label layer, and T the binary relationship between adjacent labels.
Figure 2.
Figure 2.
RaptorX Property server job submission. (A) Web interface for job submission. Job name (1) and user email address (2) are optional. Sequences in FASTA format can be submitted through the text box or a file (3). Click on the example link to see an example. Users can specify if sequence profile shall be used or not (4). Submit a job by clicking on the submit button (5). (B) An example for submission by a publicly available program Curl. Only ‘sequences’ and the submission URL (shown in underlines) are required and the others are optional. A job URL will be returned on screen after submission.
Figure 3.
Figure 3.
RaptorX Property result page. The first section shows a summary of the prediction results (1), downloading button (2), and the job status (the submitted, scheduled, and finished time) (3). The second section shows predicted results (4), with the first row showing input sequence, and the remaining rows showing the prediction of 3-state secondary structure (SS3), 8-state secondary structure (SS8), 3-state solvent accessibility (ACC), and order/disorder regions (DISO) with ‘*’ indicating disorder, respectively. The third section shows more detailed prediction results of SS3, SS8, ACC, and DISO (5), visualizing the predicted distribution at each residue (6). Hovering over a residue will display the predicted label distribution for that residue (7).

Similar articles

See all similar articles

Cited by 43 PubMed Central articles

See all "Cited by" articles


    1. Dill K.A. Dominant forces in protein folding. Biochemistry. 1990;29:7133–7155. - PubMed
    1. Myers J.K., Oas T.G. Preorganized secondary structure as an important determinant of fast protein folding. Nat. Struct. Mol. Biol. 2001;8:552–558. - PubMed
    1. Oldfield C.J., Dunker A.K. Intrinsically disordered proteins and intrinsically disordered protein regions. Annu. Rev. Biochem. 2014;83:553–584. - PubMed
    1. Bairoch A., Apweiler R. The SWISS-PROT protein sequence data bank and its supplement TrEMBL. Nucleic Acids Res. 1997;25:31–36. - PMC - PubMed
    1. Berman H.M., Westbrook J., Feng Z., Gilliland G., Bhat T., Weissig H., Shindyalov I.N., Bourne P.E. The protein data bank. Nucleic Acids Res. 2000;28:235–242. - PMC - PubMed