●I-TASSER ●I-TASSER-MTD ●C-I-TASSER ●CR-I-TASSER ●QUARK ●C-QUARK ●LOMETS ●MUSTER ●CEthreader ●SEGMER ●DeepFold ●DeepFoldRNA ●FoldDesign ●COFACTOR ●COACH ●MetaGO ●TripletGO ●IonCom ●FG-MD ●ModRefiner ●REMO ●DEMO ●DEMO-EM ●SPRING ●COTH ●Threpp ●PEPPI ●BSpred ●ANGLOR ●EDock ●BSP-SLIM ●SAXSTER ●FUpred ●ThreaDom ●ThreaDomEx ●EvoDesign ●BindProf ●BindProfX ●SSIPe ●GPCR-I-TASSER ●MAGELLAN ●ResQ ●STRUM ●DAMpred

●TM-score ●TM-align ●US-align ●MM-align ●RNA-align ●NW-align ●LS-align ●EDTSurf ●MVP ●MVP-Fit ●SPICKER ●HAAD ●PSSpred ●3DRobot ●MR-REX ●I-TASSER-MR ●SVMSEQ ●NeBcon ●ResPRE ●TripletRes ●DeepPotential ●WDL-RF ●ATPbind ●DockRMSD ●DeepMSA ●FASPR ●EM-Refiner ●GPU-I-TASSER

●BioLiP ●E. coli ●GLASS ●GPCR-HGmod ●GPCR-RD ●GPCR-EXP ●Tara-3D ●TM-fold ●DECOYS ●POTENTIAL ●RW/RWplus ●EvoEF ●HPSF ●THE-DB ●ADDRESS ●Alpaca-Antibody ●CASP7 ●CASP8 ●CASP9 ●CASP10 ●CASP11 ●CASP12 ●CASP13 ●CASP14

Methods

Starting from a query sequence, DRfold (Deep RNA fold) will use the sequence and its predicted secondary structures as the inputs of the deep learning model. The deep learning model employs 48 transformer blocks with pair representations for end-to-end RNA coordinates prediction. The pair representations are also used for inter-nucleotide geometry prediction (e.g., distance and dihedral angle maps). A novel potential that combines information from end-to-end and geometry predictions is developed to guide the L-BFGS structure optimization to obtain the final structure. (see Fig. 1).

Figure 1. Pipeline of DRfold for deep-learning based ab initio RNA structure prediction.

DRfold was tested on a set of 40 recently released targets collected from PDB. The performance of DRfold is significantly better than existing RNA structure predictors. Fig. 2 compares the performance evaluated by RMSD and TM-score of P atoms.

Figure 2. Head-to-head performance comparisons between DRfold and other RNA structure prediction methods.

The user needs to paste the RNA sequence into the input box, or upload the sequence of the query RNA using the "Choose file" button (see Fig. 3).


Figure 3. Illustration of input for the DRfold server.

The output of the DRfold server include:
• The full-length atomic model
• Predicted geometry maps from geometry models
• Predicted raw conformations from end-to-end models
An illustrative example of the DRfold output can be seen from below (Figs. 4-6):
• The full-length atomic model:
  
  

  Figure 4. Final DRfold structure in output page.

  
  

• Geometry maps:
  
  
  

  Figure 5. Geometry map visualization in output page.
  

  
  

• Predicted raw conformations from end-to-end models:
  
  

  Figure 6. Visualization of 6 raw conformation predictions of DRfold in the output page.
  

  
  

• Yang Li, Chengxin Zhang, Chenjie Feng, Robin Pearce, P. Lydia Freddolino, Yang Zhang. "Integrating end-to-end learning with deep geometrical potentials for ab initio RNA structure prediction." Nature Communications 14, no. 1 (2023): 5745.