Predicting Solute Descriptors for Organic Chemicals by a Deep Neural Network (DNN) Using Basic Chemical Structures and a Surrogate Metric

Abstract

Solute descriptors have been widely used to model chemical transfer processes through poly-parameter linear free energy relationships (pp-LFERs); however, there are still substantial difficulties in obtaining these descriptors accurately and quickly for new organic chemicals. In this research, models (PaDEL-DNN) that require only SMILES of chemicals were built to satisfactorily estimate pp-LFER descriptors using deep neural networks (DNN) and the PaDEL chemical representation. The PaDEL-DNN-estimated pp-LFER descriptors demonstrated good performance in modeling storage-lipid/water partitioning coefficient (log K storage-lipid/water ), bioconcentration factor (BCF), aqueous solubility (ESOL), and hydration free energy (freesolve). Then, assuming that the accuracy in the estimated values of widely available properties, e.g., logP (octanol−water partition coefficient), can calibrate estimates for less available but related properties, we proposed logP as a surrogate metric for evaluating the overall accuracy of the estimated pp-LFER descriptors. When using the pp-LFER descriptors to model log K storage-lipid/water , BCF, ESOL, and freesolve, we achieved around 0.1 log unit lower errors for chemicals whose estimated pp-LFER descriptors were deemed "accurate" by the surrogate metric. The interpretation of the PaDEL-DNN models revealed that, for a given test chemical, having several (around 5) "similar" chemicals in the training data set was crucial for accurate estimation while the remaining less similar training chemicals provided reasonable baseline estimates. Lastly, pp-LFER descriptors for over 2800 persistent, bioaccumulative, and toxic chemicals were reasonably estimated by combining PaDEL-DNN with the surrogate metric. Overall, the PaDEL-DNN/surrogate metric and newly estimated descriptors will greatly benefit chemical transfer modeling.

Graph Attention Network Model with Defined Applicability Domains for Screening PBT Chemicals

“…As shown in Table S11, this study employed a standardized taxonomy, ClassFire, to classify PBT chemicals identified by the traditional screening tools and descriptor-based DL models, ,, according to specific chemical classes and recognizable structural characteristics, for example, pyrimidine nucleosides, flavanols, and benzazepines. The current study further classified the PBT chemicals identified by the GAT model (3) with the loose AD FP–AC ( S cutoff = 0.80 and C cutoff = 0.70).…”

Small data machine learning in materials science

“…The acquisition of element descriptors requires the composed chemical elements of the materials and their stoichiometric ratios. Structural descriptors reflect not only compositional information, but also the 2D or 3D structural information of the materials, which can be generated by descriptor generation software or toolkits like Dragon, PaDEL, and RDkit [28][29][30][31] . Process descriptors do not reflect information about the materials themselves, but rather reflect the influence of experimental conditions in synthesis or characterization on the properties.…”

Count-Based Morgan Fingerprint: A More Efficient and Interpretable Molecular Representation in Developing Machine Learning-Based Predictive Regression Models for Water Contaminants’ Activities and Properties

“…The Morgan fingerprint of a molecule can be easily generated by the RDKit package, and the atom group in each bit can also be obtained by using the package, which facilitates model interpretation. The Morgan fingerprint has been widely used in many QSAR models. − …”