Prediction of sub-cavity binding preferences using an adaptive physicochemical structure representation

Abstract

Motivation: The ability to predict binding profiles for an arbitrary protein can significantly improve the areas of drug discovery, lead optimization and protein function prediction. At present, there are no successful algorithms capable of predicting binding profiles for novel proteins. Existing methods typically rely on manually curated templates or entire active site comparison. Consequently, they perform best when analyzing proteins sharing significant structural similarity with known proteins (i.e. proteins resulting from divergent evolution). These methods fall short when used to characterize the binding profile of a novel active site or one for which a template is not available. In contrast to previous approaches, our method characterizes the binding preferences of sub-cavities within the active site by exploiting a large set of known protein-ligand complexes. The uniqueness of our approach lies not only in the consideration of sub-cavities, but also in the more complete structural representation of these sub-cavities, their parametrization and the method by which they are compared. By only requiring local structural similarity, we are able to leverage previously unused structural information and perform binding inference for proteins that do not share significant structural similarity with known systems.

Results: Our algorithm demonstrates the ability to accurately cluster similar sub-cavities and to predict binding patterns across a diverse set of protein-ligand complexes. When applied to two high-profile drug targets, our algorithm successfully generates a binding profile that is consistent with known inhibitors. The results suggest that our algorithm should be useful in structure-based drug discovery and lead optimization.

Fig. 1.

The algorithmic framework is divided of two stages: the sub-cavity generation stage where the sub-cavities are initially identified and the sub-cavity comparison stage where sub-cavities are clustered and reshaped.

Fig. 2.

A simplified illustration of the physicochemical analysis of a sub-cavity. (A) illustrates the binding site identification described in Section 3.3. Open square, circle and triangle correspond to inner volume, binding surface and inner surface grid points, respectively. Star corresponds to a chemical group of a flanking amino acid residue. The dashed axes represent the principal axes of the sub-cavity. (B) Illustrates the initialization of the sub-cavity similarity maximization stage described in Section 3.5. The sub-cavity is rigidly transformed such that its principle axes are aligned with the x and y axes and its center of mass is at the origin. The cumulative effect of each flanking functional group (illustrated by the shaded regions) is computed for every grid point in the new coordinate frame that overlaps the sub-cavity.

Fig. 3.

Results on 20 simulated template sub-cavity structures. While the homogeneity scores are fairly consistent during refinement (A) the outlier scores dramatically improve after the first iteration (B). The percentage improvement in net cluster similarity for each noise level is shown in (C). Significant improvement is not observed for the 10% noise experiments as the clusters are already extremely similar. A larger improvement in similarity is observed for experiments run with higher noise levels.

Fig. 4.

The distribution of homogeneity scores by ligand function type for the six enzyme class experiments. The majority of clusters are highly homogeneous.

Fig. 5.

The structural scaffold common to many HIV protease inhibitors contains a central seven-member ring with seven peripheral R-groups (A). A sub-cavity is defined for each R-group. The binding site of HIV-1P 1HWR with its bound ligand is shown in (B). The extent of the binding site is illustrated as a blue mesh. A sub-cavity of one of the 1-butene groups is shown in red.