RaptorX for protein modeling and analysis
RaptorX[1][2][3] is a software and web server for protein structure and function prediction that is free for non-commercial use. RaptorX is among the most popular methods for protein structure prediction. Other popular methods include HHpredHHpred / HHsearch and Phyre2. Like other remote homology recognition/protein threading techniques, RaptorX is able to regularly generate reliable protein models when other widely used methods such as PSI-BLAST cannot. Different from profile-based methods, RaptorX excels at modeling of protein sequences without sparse sequence profiles. The RaptorX server has been designed to ensure a user-friendly interface for users inexpert in protein structure prediction methods.
Contents |
RaptorX predicts the 3D structure of a protein sequence using the principles and techniques of protein threading. Because the 3D structure of a protein is more conserved in evolution than its sequence, a protein sequence (also called target) can be modeled with reasonable accuracy on a very distantly related sequence of known structure (the template), provided that the relationship between target and template can be discerned through sequence-structure alignment. Currently almost all protein threading methods rely on sequence profiles, represented as a position-specific scoring matrix or an hidden Markov model (HMM). These profiles capture the mutational propensity of each position in a protein sequence and can be also be interpreted as an 'evolutionary fingerprint' of a particular protein. The sequence profile-based methods usually fail when the target under consideration has only a very sparse sequence profile. RaptorX not only uses sequence profile, but also heavily exploit structure information (in both single templates or multiple structurally-similar templates) for accurate protein alignment by statistical inference. Therefore, RaptorX can generate better alignments than many other methods for proteins with sparse sequence profiles.
The RaptorX project was started in 2008 and the RaptorX server was publicly released in 2011.
After pasting a protein sequence into the RaptorX "new job" submission form, a user will typically wait a couple of hours (depending on sequence length) for a prediction to complete. An email is sent to the user together with a link to a web page of results. The RaptorX server typically generates the following results: secondary structure prediction, sequence-template alignment, 3D structure prediction and functional prediction. All the result files are also available for download. The predicted 2nd and 3rd structures are displayed in color.
Applications of RaptorX include protein structure prediction, function prediction, protein sequence-structure alignment, evolutionary classification of proteins, guiding site-directed mutagenesis and solving protein crystal structures by molecular replacement. In the CASP9 blind protein structure prediction experiment, RaptorX was ranked 2nd out of about 80 automatic structure prediction servers. RaptorX also generated the best alignments for the 50 hardest CASP9 TBM(template-based modeling) targets.
RaptorX is the successor to the RAPTOR protein structure prediction system. RAPTOR was designed and developed by Dr. Jinbo Xu and Dr. Ming Li at the University of Waterloo. RaptorX was designed and developed by Mr. Jian Peng, Mr. Morten Kallberg, Dr. Haipeng Wang and Dr. Sheng Wang at the Toyota Technological Institute at Chicago. Research and development of RaptorX was supervised by Prof. Jinbo Xu.