Vikrant A. Dev

Designing Reactants and Products with Properties Dependent on Both Structures

Abstract Computer aided molecular design (CAMD) in the past has successfully been performed to design chemicals for a variety of processes including those involving reactions. However, the problem of design of reactants and products with optimal properties has not received much attention. Recently attempts have been made to address this problem. The majority of these contributions, however, are restricted to design of reactants and products when only a single reactant and product can be structurally varied. An attempt to include variation of structures of multiple reactants and products was also restricted in scope. Only the dominant properties of the product molecules could be optimized such that each product molecule was subjected to its respective set of property constraints. In this work, an algorithm has been developed that designs reactants and products such that properties that are functions of structures of reactants and products are optimized. Also, the reactants and products are subjected to their respective set of property constraints. Certain thermodynamic properties of reactions like standard Gibbs free energy change of a reaction can be optimized using the developed algorithm. Linear and nonlinear property-structure relationships based on group contributions and/or topological indices (TIs) have been utilized. They have been treated on a single platform using signature descriptors. Signatures have also been utilized to relate structures of reactants and products.

Structure Generation of Candidate Reactants Using Signature Descriptors

Abstract Integrated process and product design techniques in a reverse problem formulation are successful in selecting optimal chemicals and maximizing process efficiency. However most of these techniques are limited to non-reactive systems. Previous efforts to optimize products and generate candidate reactants have generally been restricted to a single reactant being an unknown, utilization of property prediction models that are linear functions of topological indices (TIs) and implementation of independent-site approximation among others. In this work, a molecular design algorithm has been developed that incorporates property models that are non-linear functions of TIs for the design of optimal products without restriction on number of unknown reactants and products. Property operators that are tailored functions obeying linear mixing rules and molecular signature descriptors that are novel molecular descriptors have been utilized to generate structures of reactants. Signature descriptors are capable in capturing the interactions between neighboring atoms. The algorithm can generate reactant structures of any reaction as it focuses on the design of the products and then considers the change in the various chemical bonds due to the reaction chemistry.

Reactant Structure Generation by Signature Descriptors and Real Coded Genetic Algorithm

Abstract In a reverse problem formulation, integrated process and product design techniques are successful in selecting optimal chemicals and maximizing process efficiency. However most of these techniques are limited to non-reactive systems. Initial efforts to mitigate this shortcoming have generally been restricted to a single reactant being an unknown, utilization of property models that are linear functions of topological indices (TIs) and implementation of independent site approximation among others. In this work, a molecular design algorithm has been developed that incorporates property models that are non-linear functions of TIs for the design of optimal products without restriction on the number of unknown reactants and products. For reactant structure generation, property operators that are tailored functions obeying linear mixing rules and molecular signature descriptors that are novel molecular descriptors have been utilized. Signature descriptors capably account for interactions between neighboring atoms. To tackle the ensuing nonlinear integer optimization problem(s), a real coded genetic algorithm, MI-LXPM, has been incorporated in the solution scheme. The algorithm can generate reactant structures of any reaction as it focuses on the design of the products and then considers the change in the various chemical bonds owed to the reaction chemistry.

Chapter 1 – Mathematical Principles of Chemical Product Design and Strategies

Abstract With the transformation of chemical industries from being process-focused to being product-focused, there has been remarkable progress and efforts in the field of computer-aided chemical product design. This chapter provides an overview of the various mathematical tools used for chemical product design. This chapter focuses on the utilization of mathematical programming techniques to identify/generate molecules with optimal/desirable properties. Various optimization algorithms appropriate for dealing with single and multiple objectives are described. In order to utilize such optimization techniques, a discussion of design of experiments that maximizes the collection of information is presented. The data gathered is utilized to develop property models that relate molecular structure to properties and are incorporated in the optimization procedure. A discussion of molecular descriptors, which capture structural features, is also presented. Also, the two main approaches for solving molecular design problems, i.e., the forward approach and the inverse approach, are presented. These methods are compared to the traditional product design approach, which relies primarily on experiments. The consideration of uncertainty in the computer-aided design procedures is also discussed in this chapter. Finally, further development possibilities in the field of chemical product design are discussed.

Multi-Objective Computer-Aided Molecular Design of Reactants and Products

Abstract Most chemical products are obtained from processes that involve reactions. Although the problem of designing solvents and catalysts for such reactive systems has garnered considerable attention, focus on the problem of design of reactants and products has been lacking. Recently an attempt was made to design multiple reactants and products such that properties dependent on structures of both reactants and products are optimized. However, only the optimization of a single objective was studied. In this work, an algorithm is being presented to design reactants and products such that their respective dominant properties are optimized. Also each of the reactants and products are subjected to a set of property constraints. The algorithm also allows for the inclusion of properties dependent on structures of both reactants and products as an objective. In this algorithm, we have utilized signature descriptors, which are molecular building blocks. Both linear and nonlinear structure-property relationships, expressed in terms of group contributions (GCs) and/or topological indices (TIs), can be utilized in this algorithm. Signature descriptors help treat these models on a single platform. To illustrate the efficacy of our algorithm, a case study is presented in this paper.

Hybrid genetic algorithm-decision tree approach for rate constant prediction using structures of reactants and solvent for Diels-Alder reaction

Abstract In recent years, Computer-Aided Molecular Design (CAMD) has been extensively used for defining and designing reactions at their maximal potential. In all of these contributions, either the structures of reactants/products have been considered to be unchanging or the solvent structure. Developing a QSPR model which not only captures the influence of reactant structures but also the solvent effect on reaction rate, is essential. Since the structures of reactants and products are related, such QSPR models will serve as a prerequisite for the simultaneous CAMD of reactants, products and solvents. They will also provide a useful tool for predicting the rate constant without relying on experiments. To develop such a QSPR, in our work, the Diels-Alder reaction with different sets of reactants and solvents was investigated. Connectivity indices were used to represent the structures of the members of each set. Principal Component Analysis (PCA) was applied to identify principal components (PCs) corresponding to the structures of reactants and solvent of each set. Linear models expressed in terms of PCs were then generated using a Decision Tree (DT) algorithm such that the R 2 value was maximized. These models formed the initial population on which the GA performed operations such as crossover and mutation to obtain model(s) with best rate constant prediction. Thus, the novelty of our approach is that after feature extraction using PCA, a DT algorithm generates an ensemble of linear models, which through the GA is transformed into a model with best fit. Our approach required much lesser generations to provide a model with highest R 2 ext value as compared to the case where the DT did not initialize the population of models.

Developing QSPR for Predicting DNA Drug Binding Affinity of 9-Anilinoacridine Derivatives Using Correlation-Based Adaptive LASSO Algorithm

Abstract Major advancements in the field of machine learning and readily available inexpensive computational power, QSPRs (quantitative structure property relationships) are increasingly being viewed, by the scientific community, as reliable tools that can provide accurate property prediction. Additionally, QSPRs offer advantages of experimental cost reduction and reduction in chemical footprint associated with experiments. Treatment of cancerous tumors has become a global focus due to the heightened prevalence of such tumors in humans, both young and old. Apart from surgery, the most commonly used treatment is chemotherapy. As there are many long-term side effects of chemotherapy such as organ damage, fatigue, hair loss and tooth loss, researchers are devoting much attention to the search of treatments with fewer side effects. So far, no effective solution has emerged which can be reported as an alternative to chemotherapy. In a recent study, thirty one (31) 9-anilinoacridines were synthesized and evaluated for their antitumor activity. The association constant, K , was utilized as a key determining factor to evaluate the DNA drug binding affinity. 9-anilinoacridines show great promise as antitumor agents. In order to help reduce the experimental effort of K value determination and to assist in the design of 9-anilinoacridines, in this work, we developed a QSPR to predict K . In order to develop the QSPR, all the structures were drawn and optimized using the Avogadro software and converted to mol files. The Dragon 6 software was then used to calculate the values of descriptors using the generated mol files. The descriptors were then used to develop the model using GA (genetic algorithm) and CorrLASSO (correlation-based adaptive least absolute shrinkage and selection operator). The CorrLASSO in combination with GA helped generate a model with superior prediction as compared with the combination of GA and LASSO (least absolute shrinkage and selection operator) and GA-MLR (genetic algorithm-multiple linear regression). In our work, R 2 , Q 2 and MSE (mean squared error) calculations have been performed to assess model performance and data fitness.

Comparison of Tree Based Ensemble Machine Learning Methods for Prediction of Rate Constant of Diels-Alder Reaction

Abstract The design of molecular solvents has garnered significant interest because solvents have been shown to influence the rate of chemical product generation in a reaction. In order to quantitatively understand the influence of solvent structure on the rate of the reaction, models are needed that capture this influence, in addition to that of the reactants’ structure, on the rate constant. A quantitative structure-property relationship (QSPR) for the Diels-Alder reaction was recently developed using a hybrid genetic algorithm-decision tree (GA-DT) approach. However, there is still scope for improvement in the performance of the QSPR. In an attempt to further improve upon the performance of the aforementioned QSPR, we have assessed various tree based ensemble machine learning regression methods for prediction of rate constant (modeled using connectivity indices) of Diels-Alder reaction. The assessed methods are random forest regression, gradient boosted regression trees, regularized random forest regression and extremely randomized trees. The evaluation was carried out in terms of the R2 and Q2 values. Extremely randomized trees were found to provide the highest R2 value of 0.91 while random forests provided the highest Q2 value of 0.76.

Relating Reaction Rate Constant to Structures of Reactants and Solvent Using a Hybrid GA-DT Approach

Abstract In recent years, Computer-Aided Molecular Design (CAMD) has been extensively used for defining and designing reactions at their maximal potential. In all these contributions, either the reactants/products were considered constant or the solvents. Developing a QSPR model which not only captures the influence of reactant structures but also the solvent effect on reaction rate, is essential. Since the structures of reactants and products are related, such QSPR models will serve as a prerequisite for the simultaneous CAMD of reactants, products and solvents. They will also provide a useful tool for predicting the rate constant without relying on experiments. To develop such a QSPR, in our work, the Diels-Alder reaction with different sets of reactants and solvents was investigated. Connectivity indices were used to represent the structures of the members of each set. Principal Component Analysis (PCA) was applied to identify the principal components (PCs) of each set for further use in model development. These PCs were then used to develop a linear model that best predicts the reaction rates in our study. In this paper, Genetic Algorithm (GA) has been modified using the Decision Tree (DT) algorithm for increased efficiency. Inclusion of DT in GA ensures an initial generation of meaningful combination of descriptors. This set gets further improved in every step of crossover and mutation with applied constraints. Only improvement of generations is accepted due to these constraints. Finally, Multiple Linear Regression (MLR) relates the chosen descriptors with the property under study. The model undergoes thorough internal and external validation to ensure that a best fit model can be found in minimum steps possible.