Rosetta: De Novo determination of protein structure www.bakerlab.org
Results and performance Application and Future Prospects Conclusion Outline Background- Ab Initio predictions Rosetta Method Results and performance Application and Future Prospects Conclusion
Simulate interactions between pairs of atoms Background Ab Initio predictions Physics of atomic interactions Simulate interactions between pairs of atoms Very difficult Heuristic Methods Based on databses of known structures
Background Homology modeling Similarity to known structure Easy if sequence similarity > 50% Many methods, still popular Sequence-structure bias key to predict structure from sequence?
Beg, Borrow or Steal - Structures The Folding Problem Search Problem Beg, Borrow or Steal - Structures Discrimination Problem Identification of native-like conformation
Venn diagram of conceptual basis for Rosetta Rosetta Method Natural folding process Local structure propensities Tertiary Structure Local Tertiary N Venn diagram of conceptual basis for Rosetta
Secondary Structure Prediction The Rosetta Protocol Sequence Secondary Structure Prediction 3 and 9 residues fragment selection Monte-Carlo fragment insertion Clustering and Filtering of decoys Full-atom refinement 3D Structure
Multiple Sequence Alignment Consistent fragments of 2 lengths Protocol Contd. Fragment Generation Secondary Structure Prediction PsiPred , PHD , DSC and JUFO + Multiple Sequence Alignment Consistent fragments of 2 lengths
Polar Side Chain Interactions Hydrogen Bonding – Beta Strands Scoring Functions Hydrophobic Burial Polar Side Chain Interactions Hydrogen Bonding – Beta Strands Hard Sphere Repulsion
Clustering and Filtering of Decoys Global Backbone RMSD 80-100 Structure in largest Cluster Contact Order Cut-off Radius of Gyration Cut-off Empirical Formula Beta Strand Pairing
Results Native Native Model Native Model Model
More Results…
Folding Pathway of Ubiquitin Helix Formation – 12 fragments Beta hairpin – 460 fragments Native Fold – 760 fragments Final refinement – 1304 frags RMSD ~ 2.1 A www.Jens-meiler.de
Size of the proteins ~ 100 residues Success and Failures Native Like Conformation ( 4-6 A ) are produced CASP results Size of the proteins ~ 100 residues Difficult topologies still not good Computationally expensive
Future Prospectus and Applications RosettaNMR, RosettaNOE With Minimal Experimental Constraints RosettaNMR, RosettaNOE 3D Structure without Resonance Assignments ITAS, MJ’s Program Integration with other Ab Initio programs HMMSTR and I-Sites
Conclusion Low to medium resolution structure are possible using Rosetta method. These structure are good starting point for further refinement. Low Resolution structure can give insight to protein functions.
Thank You!!