Protein Structure Property Prediction
RaptorX Property is a web server predicting structure property of a protein sequence without using any template information. It outperforms other servers especially for proteins without close homologs in the Protein Data Bank (PDB) or with very sparse sequence profile. This server employs an emerging machine learning model called DeepCNF (Deep Convolutional Neural Fields) to predict secondary structure (SS), solvent accessibility (ACC), and disorder regions (DISO) simultaneously. 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 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. RaptorX-Property was ranked 1st in 3-/8-state secondary structure prediction in a third-party evaluation work published in Briefings in Bioinformatics. Software is available here.