Rong Chen Boston University Protein Docking Rong Chen Boston University
The Lowest Binding Free Energy DG water L R L R R L R L L R
Protein Docking Using FFT Fast Fourier Transform Complex Conjugate Discretize Correlation function Fast Fourier Transform L L Rotate L Discretize Surface Interior
Rotational Sampling Sampling Interval Number of angles 20° 1,800 15° Evenly distributed Euler angles Sampling Interval Number of angles 20° 1,800 15° 3,600 12° 9,000 10° 14,400 8° 27,000 6° 54,000 4° 180,000
Performance Evaluation Success Rate: given the number of predictions(Np), success rate is the percentage of complexes in the benchmark for which at least one hit has been obtained. Hit Count: the average number of hits over all complexes at a particular Np.
Rotational Sampling Density
Rotational Sampling Density
Protein Docking Using FFT Fast Fourier Transform Complex Conjugate Discretize Correlation function Fast Fourier Transform L L Rotate L Discretize Surface Interior
Protein Docking Using FFT Increase the speed by 107 Y Translation Correlation X Translation L R IFFT Surface Interior Binding Site
An Effective Binding Free Energy Function van der Waals energy; Shape complementarity Desolvation energy; Hydrophobicity Electrostatic interaction energy Translational, rotational and vibrational free energy changes Number of atom pairs of type i Desolvation energy for an atom pair of type i
Grid-based Shape Complementarity 1 RGSC LGSC
Pairwise Shape Complementarity 2 3 5 1 RPSC LPSC
PSC vs. GSC on Success Rate
PSC vs. GSC on Hit Count
Why PSC works better than GSC?
Why PSC works better than GSC? D
A Receptor-Ligand Complex
An Effective Binding Free Energy Function van der Waals energy; Shape complementarity Desolvation energy; Hydrophobicity Electrostatic interaction energy Translational, rotational and vibrational free energy changes Number of atom pairs of type i-j Desolvation energy for an atom pair of type i-j
Impact of Desolvation and Electrostatics
Impact of Desolvation and Electrostatics
Other available Docking Software Fast Fourier Transform or FFT (Katchalski-Katzir, Sternberg, Vakser, Ten Eyck groups) Computer vision based method (Nussinov group, 1999) Boolean operations (Palma et al., 2000) Polar Fourier correlations (Ritchie & Kemp, 2000) Genetic algorithm (Gardiner, Burnett groups) Flexible docking (Abagyan, 2002)
3D-Dock Michael J.E. Sternberg, Imperial Cancer Research Fund, London, UK. FTDock: Grid-based shape complementarity, FFT. RPScore: empirical pair potential. MultiDock: refinement. http://www.bmm.icnet.uk/docking/index.html
GRAMM Ilya A. Vakser, State University of New York at Stony Brook. Geometric fit and hydrophobicity FFT Low resolution docking http://reco3.ams.sunysb.edu/gramm/
DOT Lynn F. Ten Eyck, University of California, San Diego. Grid-based shape complemetarity, elctrostatics FFT http://www.sdsc.edu/CCMS/Papers/DOT_sc95.html
ICM Ruben Abagyan, The Scripps Research Institute, La Jolla. Pseudo-Brownian rigid-body docking Biased Probability Monte Carlo Minimization of the ligand interacting side-chains. http://abagyan.scripps.edu/lab/web/man/frames.htm
HEX Dave Ritchie, University of Aberdeen, Aberdeen, Scotland, UK spherical polar Fourier correlations http://www.biochem.abdn.ac.uk/hex/
Approach Overview PDB1 PDB2 PDB Processing Biological information ZDOCK: Initial-stage Docking Biological information RDOCK: Refinement-stage Docking Clustering Final 10 predictions
Example: CAPRI Target 6: α-amylase / Camelid VHH domain