3rd Permanent School in Bioinformatics Madrid 2005 Protein Structure Modeling

1. 3rd Permanent School in Bioinformatics Madrid 2005 Protein Structure Modeling Alejandro Giorgetti Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

2. Public Database Holdings: The number of different protein folds is limited: Known Folds [ last update: Oct 2001 ] NewFolds Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

3. Fold recognition Q M T S A F G T A E Profile methods • Fold =f(environment) • Local Secondary Structure • Solvent Accessibility • Degree of burial of polar / apolar Principle: Find a compatible fold >Target Sequence XY MSTLYEKLGGTTAVDLAVAAVAGAPAHKRDVLNQ Threading Fragment based methods: New fold prediction Build model of target protein based on each template structure Rank models according to SCORE or ENERGY Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

4. Some remarks concerning fold recognition: • Capable of detecting quite remote sequence-structure matches. • Sensitivity depends on the size of the protein and its secondary structure content. • The two most versatile enzymatic functions (hydrolases and o-glycosyl- glucosidases) are associated with seven folds each. • Better for detecting: all-α > αβ > all-β Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

5. Homology modeling • Comparative protein modeling Idea:Proteins evolving from a common ancestor maintained similar core 3D structures. • Known structure/s is/are used as a template to model an unknown structure with known sequence. • Both of them should be related by evolution. • First applied in late 1970’s by Tom Blundell Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

6. Evolution of protein structure families 90 % Drug design? 70 % Biochemistry? X-ray cristallography: MR 50 % 30 % Molecular Biology? 10 % [ Chothia & Lesk (1986) ] Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

7. Comparative Modeling Known Structures (templates) Template(s) selection Target sequence Sequence Alignment Structure Evaluation >hTEII MSSPQAPEDGQGCGDRGDPPGDLRSVLVTTV LNLEPLDEDLFRGRHYWVPAKRLFGGQIVGQ ALVAAAKSVSEDVHVHSLHCYFVRAGDPKLP Structure Modeling Final Structural Models Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

8. Comparative Modeling Known Structures (templates) Template(s) selection Target sequence • Protein Data Bank PDB http://www.pdb.org • Database of templates • Separate into single chains • Remove bad structures (models) • Create BLAST database Sequence Alignment Structure Evaluation Structure Modeling Final Structural Models Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

9. Comparative Modeling Known Structures (templates) Template(s) selection Target sequence • Sequence Similarity / Fold recognition • Structure quality (resolution, experimental method) • Experimental conditions (ligands and cofactors) Sequence Alignment Structure Evaluation Structure Modeling Final Structural Models Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

10. Comparative Modeling Known Structures (templates) Template(s) selection Target sequence • Key step in homology modeling • Global alignment is required • Small error in alignment can lead to big error in model • Multiple alignments are better than pairwise alignments • Do we know something else? Experiments? Sequence Alignment Structure Evaluation Structure Modeling Final Structural Models Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

11. Comparative Modeling Known Structures (templates) Template(s) selection Target sequence Sequence Alignment Structure Evaluation • Template based fragment Assembly (SwissMod). • Satisfaction of Spatial Restraints: MODELLER Structure Modeling Final Structural Models Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

12. Comparative Modeling Known Structures (templates) Template(s) selection Target sequence • Errors in template selection or alignment result in bad models • Iterative cycles of alignment, modeling and evaluation Sequence Alignment Structure Evaluation Structure Modeling Final Structural Models Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

13. I. Template based fragment assembly (SwissModel) [ http://www.expasy.org/spdbv/ ] Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

14. Day activities. Morning • SwissPdb downloading: • Read and Accept licence • Download: SwissPdb viewer v3.7sp5(linux) • Installation: • gunzip spdbv37sp5-Linux.tar.gz • tar –xvf spdbv37sp5-Linux.tar • cd SPDBV_Distribution and do: ./install.sh • Local installation : ..../guestxx/ • Run: /guestxx/SPDBV/bin/spdbv • SwissModel submission: • Search template: http://www.expasy.org/swissmod/SWISS-MODEL.html Interactive tools: Search the template...(paste sequence) • Model request submission: Save project (SwissPdb Viewer); • Swiss- Model web page: Modelling Requests – Optimise mode. • Fill the form and Upload your project asking for Short mode output. Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

15. I. Template based fragment assembly a) Build conserved core framework (Structurally conserved regions -SCRs) [ http://www.expasy.org/spdbv/ ] • Corresponds to the most stable regions. • Highest sequence conservation and fewer gaps. • In general: secondary structures elements. Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

16. I. Template based fragment assembly b) Loop modeling (Structural variable regions - SVRs) and backbone completion • Least stable or more flexible regions. • Highest level of gapping • Lowest sequence conservation • Loops and turns • Loop-Database • “ab-initio” rebuilding of loops (Monte Carlo, molecular dynamics, genetic algorithms, etc.) [ http://www.expasy.org/spdbv/ ] Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

17. I. Template based fragment assembly c) Side Chain placement • Find the most probable side chain conformation, using • homologues structures • back-bone dependent rotamer libraries • energetic and packing criteria [ http://www.expasy.org/spdbv/ ] Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

18. I. Template based fragment assembly d) Energy minimization • modeling will produce unfavorable contacts and bonds •  idealization of local bond and angle geometry • extensive energy minimization will move coordinates away •  keep it to a minimum • SwissModel is using GROMOS 96 force field Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

19. II. Modeling by Satisfaction of Spatial restraints • Find the most probable structure given its alignment • Satisfy spatial restraints derived from the alignment. • Uses probability density functions. • Minimizes violations on restraints. Comparative protein modeling by satisfaction of spatial restraints. A. Šali and T.L. Blundell. J. Mol. Biol. 234, 779-815 Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

20. Model Evaluation ? • Topics: • correct fold • model coverage (%) • C - deviation (rmsd) • alignment accuracy (%) • side chain placement • Structure Analysis and Verification Server: http://nihserver.mbi.ucla.edu/SAVS/ Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

21. EVA Evaluation of Automatic protein structure prediction [ Burkhard Rost, Andrej Sali, http://maple.bioc.columbia.edu/eva ] Model Accuracy Evaluation CASP Community Wide Experiment on the Critical Assessment of Techniques for Protein Structure Prediction http://PredictionCenter.llnl.gov/casp6 3D - Crunch Very Large Scale Protein Modelling Project http://www.expasy.org/swissmod/SM_LikelyPrecision.html Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

22. Protein Structure Resources PDB http://www.pdb.orgPDB – Protein Data Bank of experimentally solved structures (RCSB) CATH http://www.biochem.ucl.ac.uk/bsm/cath Hierarchical classification of protein domain structures SCOP http://scop.mrc-lmb.cam.ac.uk/scopAlexey Murzin’s Structural Classification of proteins DALI http://www2.ebi.ac.uk/daliLisa Holm and Chris Sander’s protein structure comparison server SS-Prediction and Fold Recognition PHD http://cubic.bioc.columbia.edu/predictprotein Burkhard Rost’s Secondary Structure and Solvent Accessibility Prediction Server PSIPRED http://bioinf.cs.ucl.ac.uk/psipred/ L.J McGuffin, K Bryson & David T. Jones Secndary struture prediction Server 3DPSSM http://www.sbg.bio.ic.ac.uk/~3dpssFold Recognition Server using 1D and 3D Sequence Profiles coupled. THREADER: http://bioinf.cs.ucl.ac.uk/threader/threader.html David T. Jones threading program Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

23. Protein Structure Classification CATH - Protein Structure Classification [ http://www.biochem.ucl.ac.uk/bsm/cath_new/ ] • UCL, Janet Thornton & Christine Orengo • Class (C), Architecture(A), Topology(T), Homologous superfamily (H) SCOP - Structural Classification of Proteins • MRC Cambridge (UK), Alexey Murzin, Brenner S. E., Hubbard T., Chothia C. • created by manual inspection • comprehensive description of the structural and evolutionary relationships [ http://scop.mrc-lmb.cam.ac.uk/scop/ ] Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

24. Class(C)derived from secondary structure content is assigned automatically • Architecture(A)describes the gross orientation of secondary structures, independent of connectivity. • Topology(T) clusters structures according to their topological connections and numbers of secondary structures • Homologous superfamily (H) Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

25. Protein Homology Modeling Resources SWISS MODEL: http://www.expasy.org/swissmod/SWISS-MODEL.html Deep View - SPDBV: homepage: http://www.expasy.ch/spdbv Tutorials http://www.expasy.org/spdbv/text/tutorial.htm WhatIf http://www.cmbi.kun.nl:1100/ Gert Vriend’s protein structure modeling analysis program WhatIf Modeller: http://guitar.rockefeller.edu/modeller Andrej Sali's homology protein structure modelling by satisfaction of spatial restraints ROBETTA: http://robetta.bakerlab.org/ Full-chain Protein Structure Prediction Server Programs and www servers very useful in Comparative modeling: http://salilab.org/tools/ Alejandro Giorgetti alejandro.giorgetti@uniroma1.it

26. Alejandro Giorgetti alejandro.giorgetti@uniroma1.it