Hierarchical PLS Modeling for Predicting the Binding of a Comprehensive Set of Structurally Diverse Protein−Ligand Complexes

Abstract

A new approach is presented for predicting ligand binding to proteins using hierarchical partial-least-squares regression to latent structures (Hi-PLS). Models were based on information from the 2002 release of the PDBbind database containing (after in-house refinement) high-resolution X-ray crystallography and binding affinity (Kd or Ki) data for 612 protein-ligand complexes. The complexes were characterized by four different descriptor blocks: three-dimensional (3D) structural descriptors of the proteins, protein-ligand interactions according to the Validate scoring function, binding site surface areas, and ligand 2D and 3D descriptors. These descriptor blocks were used in Hi-PLS models, generated using both linear and nonlinear terms, to relate the characterizations to pKd/i. The results show that each of the four descriptor blocks contributed to the model, and the predictions of pKd/i of the internal test set gave a root-mean-square error of prediction (RMSEP) of 1.65. The data were further divided according to the structural classification of the proteins, and Hi-PLS models were constructed for the resulting subclasses. The models for the four subclasses differed considerably in terms of both their ability to predict pKd/i (with RMSEPs ranging from 0.8 to 1.56) and the descriptor block that had the strongest influence. The models were validated with an external test set of 174 complexes from the 2003 release of the PDBbind database. The overall results show that the presented Hi-PLS methodology could facilitate the difficult task of predicting binding affinity.

A novel method for protein‐ligand binding affinity prediction and the related descriptors exploration

“…38,41 To study the comprehensive information of PLCs, the whole refined set was used in this work without any additional filter of PLCs, like other works. 34 Within the 1300 PLC data in the refined set, there are 493 data with the binding affinity of dissociation constant (K d ) value and 807 with inhibition constant (K i ) value. We used the negative logarithm of K d and K i values in this study (pK d and pK i ).…”

“…The external validation results are better than that in the references based on the same kind of dataset. 16,34,39 Besides, in order to further prove the generalization ability of our model, an overall 10-fold crossvalidation of this model on the whole datasets was also performed. Because the Q 2 result is not stable for an n-fold crossvalidation, we repeat this procedure for 10 times and get an average Q 2 for K d and K i datasets as 0.569 and 0.496.…”

“…Recently, as an alternative to widely used docking and scoring approach, some other in silico methods based on the structures of ligands and the relevant proteins are also proposed for the fast prediction of the binding affinity with some success (e.g. Hi-PLS 34 and novel geometrical descriptors-based method 35,36 ). These methods often use the molecular descriptors calculated from the structure of the ligand and protein as the inputs, and then, use some modeling methods to develop predictive models for binding affinity.…”

“…Many of the methods mentioned earlier used the refined set [2003 release, containing 800 protein-ligand complexes (PLCs)] of PDBbind, 37,38 including the methods of PMF, ChemScore, PLP, LUDI, GOLD, and X-Score, Hi-PLS, etc. 16,34,39 Second, some prediction models gave very good accuracy, but they were based on only a relatively small data set. For instance, the work by Zhang et al 35 used a data set containing 264 PLCs with binding affinities (pK d ) and yielding the best R 2 ext of the models as 0.83.…”

In silico structure‐activity‐relationship (SAR) models from machine learning: a review

“…In this case, the Chemogenomics-based data representation requires three ingredients of efforts. The first is the features for the targets (e.g., protein structures [Lindströ m et al, 2006], amino acid sequence [Jacob and Vert, 2008], binding site descriptors [Strö mbergsson et al, 2008;Deng et al, 2004], etc.). This part is new from conventional single-target oriented SAR methods, but has already been well studied in structural biology independently, and thus in principle a good feature representation from their studies can be applied directly for chemogenomics.…”