Machine-learning-assisted molecular design of phenylnaphthylamine-type antioxidants

Abstract

In this study, a total of 302 molecular structures of phenylnaphthylamine antioxidants based on N-phenyl-1-naphthylamine and N-phenyl-2-naphthylamine skeletons with various substituents were modeled by exhaustive methods. Antioxidant parameters, including the...

Dynamics study of the post-transition-state-bifurcation process of the (HCOOH)H+ → CO + H3O+/HCO+ + H2O dissociation: application of machine-learning techniques

“…The accuracy of each classification model was obtained by performing a 5-fold cross validation on the most relevant 54 TS features. To obtain the accuracy scores with eliminating the dimension influences of each feature, 57 the data sets were normalized into [0, 1]. An optimized RF (RF Opt) and LGB models were obtained by performing hyperparameter optimization using the GridSearchCV method of the Scikit-learn library in Python3.…”

“…The ML classification approach helps us to identify the significant properties that dominate the bifurcation dynamics without relying on human intuition. 57 Similar ML analysis techniques have been applied to other chemical reaction systems that exhibit PTSB. [58][59][60] The details of the D-ML method, the MD method and those of the supervised ML algorithm are described in Section 2.…”

The Versatile Biocatalyst of Cytochrome P450 CYP102A1: Structure, Function, and Engineering

“…Through two cycles of training and testing, a 20-fold increase in proteinase K activity was achieved after testing only 95 variants. Du and coworkers used ML to assist in the design of new antioxidants [ 61 ]. Han et al reported an ML application to improve protein solubility and activity [ 62 ].…”

Automatically Predicting Material Properties with Microscopic Images: Polymer Miscibility as an Example

“…These blends have been extensively researched due to their capacity to combine the advantageous properties of different polymers and their ease of synthesis and processing. Researchers have utilized machine learning methods to investigate their properties, including the mechanical properties, liquid crystal behavior, thermal conductivity, dielectric constants, optical properties, and molecular design . Among all kinds of properties, polymer miscibility is the primary basis for determining the structure and properties of blends, and it is difficult for immiscible polymers to form materials with good properties.…”

“…Researchers have utilized machine learning methods to investigate their properties, including the mechanical properties, 14 liquid crystal behavior, 15 thermal conductivity, 16 dielectric constants, 17 optical properties, 18 and molecular design. 19 Among all kinds of properties, polymer miscibility is the primary basis for determining the structure and properties of blends, 20 determining whether different polymers are miscible is a crucial question and attracts researchers' focus. 21 While the miscibility of some blends needs to be determined through technologies in physical chemistry such as differential scanning calorimetry (DSC), microscopic observation (such as SEM) is the most efficient approach since this property can be manifested in the homogeneity of the morphological appearance at most times.…”