Complementary PLS and KNN algorithms for improved 3D-QSDAR consensus modeling of AhR binding

Svetoslav H Slavov,Richard D Beger,Dan A Buzatu,Bruce A Pearce,Jon G Wilkes

doi:10.1186/1758-2946-5-47

Abstract

Multiple validation techniques (Y-scrambling, complete training/test set randomization, determination of the dependence of R2test on the number of randomization cycles, etc.) aimed to improve the reliability of the modeling process were utilized and their effect on the statistical parameters of the models was evaluated. A consensus partial least squares (PLS)-similarity based k-nearest neighbors (KNN) model utilizing 3D-SDAR (three dimensional spectral data-activity relationship) fingerprint descriptors for prediction of the log(1/EC50) values of a dataset of 94 aryl hydrocarbon receptor binders was developed. This consensus model was constructed from a PLS model utilizing 10 ppm x 10 ppm x 0.5 Å bins and 7 latent variables (R2test of 0.617), and a KNN model using 2 ppm x 2 ppm x 0.5 Å bins and 6 neighbors (R2test of 0.622). Compared to individual models, improvement in predictive performance of approximately 10.5% (R2test of 0.685) was observed. Further experiments indicated that this improvement is likely an outcome of the complementarity of the information contained in 3D-SDAR matrices of different granularity. For similarly sized data sets of Aryl hydrocarbon (AhR) binders the consensus KNN and PLS models compare favorably to earlier reports. The ability of 3D-QSDAR (three dimensional quantitative spectral data-activity relationship) to provide structural interpretation was illustrated by a projection of the most frequently occurring bins on the standard coordinate space, thus allowing identification of structural features related to toxicity.

Highlights

During the past decade, the application of consensus modeling to various Quantitative Structure - Activity relationship (QSAR) related problems has been explored [1,2,3]
Because T is calculated from row vectors, it can be demonstrated that k-nearest neighbors (KNN) operating on T may capture information complementary to that of the respective partial least squares (PLS) models
The predictive power of the PLS and KNN models in terms of R2test as a function of the granularity of the ThreeDimensional Spectral Data - Activity Relationship (3D-SDAR) abstract space is shown on Figure 4

Summary

Introduction

The application of consensus modeling to various QSAR related problems has been explored [1,2,3]. QSARs often relied on single models, which under certain circumstances were prone to arbitrary overestimation of the contribution of given structural features at the expense of others that were suppressed or ignored. To mitigate such risks consensus models based on a multitude of individual models can be advantageously used. 3D-QSDAR utilizes unique fingerprints constructed from pairs of 13C chemical shifts augmented with their corresponding inter-atomic distances. In our earlier work [8] an automated partial least squares (PLS) algorithm was used to process data from regularly tessellated 3D-SDAR fingerprints and to derive averaged (composite model) predictions from 100 randomized training/hold-out test set pairs. A Y-scrambling procedure [11,12] assessed the probability of generating seemingly “good” models by chance

Methods

Results

Conclusion