Advancing molecular graphs with descriptors for the prediction of chemical reaction yields

Abstract

Chemical yield is the percentage of the reactants converted to the desired products. Chemists use predictive algorithms to select high‐yielding reactions and score synthesis routes, saving time and reagents. This study suggests a novel graph neural network architecture for chemical yield prediction. The network combines structural information about participants of the transformation as well as molecular and reaction‐level descriptors. It works with incomplete chemical reactions and generates reactants‐product atom mapping. We show that the network benefits from advanced information by comparing it with several machine learning models and molecular representations. Models included logistic regression, support vector machine, CatBoost, and Bidirectional Encoder Representations from Transformers. Molecular representations included extended‐connectivity fingerprints, Morgan fingerprints, SMILESVec embeddings, and textual. Classification and regression objectives were assessed for each model and feature set. The goal of each classification model was to separate zero‐ and non‐zero‐yielding reactions. The models were trained and evaluated on a proprietary dataset of 10 reaction types. Also, the models were benchmarked on two public single reaction type datasets. The study was supplemented with analysis of data, results, and errors, as well as the impact of steric factors, side reactions, isolation, and purification efficiency. The supplementary code is available at https://github.com/SoftServeInc/yield-paper.

Exploring Chemical Reaction Space with Machine Learning Models: Representation and Feature Perspective

“…In addition to product prediction and retrosynthesis where the essence is to define and model the changes in the local structure, works have shown that applying representations based on molecular graphs of reactants, reagents, and the products could be beneficial for more reaction tasks, e.g., yield prediction − and condition prediction. , For example, Kwon et al used GNN to embed the reaction into the latent space and then apply variational inference to obtain multiple reaction conditions from the GNN encoded space. Though implementing graphs to represent reactions possesses strength, molecular graphs still have limitations, as they lack critical information such as charges, energies, and steric effects.…”

Comment on ‘Physics-based representations for machine learning properties of chemical reactions’

“…If we are interested in assessing model performance on new molecules, we can train a model with many reaction templates but use substructure splitting to create training, validation, and testing sets. Bemis-Murcko scaffolds [70] are commonly used to partition the data for this purpose, though clustering based on other input features or chemical similarity to measure extrapolation has also been explored [23,[71][72][73][74][75][76][77][78][79][80][81][82][83][84][85][86][87][88] as has quantifying domains of model applicability [89][90][91][92][93]. Scaffold splitting is not perfect, but by ensuring that molecules in the testing set are structurally different than those in the training set, it offers a much better assessment of generalizability than splitting randomly [17,24,67,[94][95][96][97][98][99][100][101][102][103][104][105][106][107][108][109]…”

When Yield Prediction Does Not Yield Prediction: An Overview of the Current Challenges

“…Yarish et al 91 developed the directed message-passing neural network (RD-MPNN) yield prediction models, which they tested on Enamine's proprietary reaction data. Their binary classification model showed a commendable ROC AUC of 0.78.…”