Task 2 : DockTrina – Docking Triangular Trimers


DockTrina is a novel protein docking method for modeling the 3D structures of nonsymmetrical triangular trimers. The method takes as input pair-wise contact predictions from a rigid body docking program. It then scans and scores all possible combinations of pairs of monomers using a very fast root mean square deviation (RMSD) test (see below). Finally, it ranks the predictions using a scoring function which combines triples of pair-wise contact terms and a geometric clash penalty term. The overall approach takes less than 2 min per complex on a modern desktop computer.

Figure 1. Workflow of the DockTrina algorithm.


The DockTrina algorithm is based on fast computation of RMSD. While scanning all possible combinations of pair-wise docking transforms, a tetramer of chains A,B,C and A′ is formed. Here A′ is an image of A, obtained with the net transform TAA:

$$T^{AA}=T^{AB}\cdot T^{BC} \cdot T^{CA},$$$$
A′=T^{AA}\cdot A$$
If the individual pair-wise transforms are perfectly mutually consistent, then chain A should be transformed into itself. It means that RMSD between chains A and A′ characterizes the quality of a trimeric prediction. We use the RigidRMSD library to perform each RMSD test in constant time, which makes DockTrina fast and efficient tool to predict triangular trimers.

Figure 2. A tetramer of chains A,B,C and A′.

Benchmark and Results

To test and validate the DockTrina algorithm, we have collected two trimer docking benchmarks, consisting of 220 bound and seven unbound protein complexes, respectively. The bound benchmark includes 85 proteins with crystallographic symmetry, 76 protein complexes related by noncrystallographic symmetry (NCS), and 59 nonsymmetrical protein complexes. The unbound benchmark consists of seven complexes, three of  which possess threefold crystallographic symmetry. On the bound docking benchmark, our algorithm obtains 66.4% acceptable predictions listed in the top 10, 55.5% near-native predictions listed in the top 10, 52.7% acceptable predictions ranked 1, and 35.9% near-native predictions ranked 1. Given that Hex did not produce any acceptable pair-wise contacts for 50 cases, and any near-native contacts for 66 cases, the suc- cess rates of our algorithm are 85.9, 79.2, 68.2, and 79.2%, respectively. We find that the performance of DockTrina on symmetrical proteins is similar to that of SymmDock algorithm, which was specifically developed to deal with cyclic symmetries. However, DockTrina gives significantly better results than HADDOCK and Comb- Dock on protein trimers without symmetry.

Figure 3. Some examples of bound complexes from our trimer benchmark set. Top row: symmetrical structures. Middle row: NCS structures. Bottom row: nonsymmetrical structures. 



P. Popov, D.W. Ritchie, and S. Grudinin. DockTrina: Docking triangular protein trimers. Proteins, 2013.  Full Text.