Abstract:In drug discovery, knowledge of the graph structure of chemical compounds is essential. Many thousands of scientific articles and patents in chemistry and pharmaceutical sciences have investigated chemical compounds, but in many cases, the details of the structure of these chemical compounds are published only as an image. A tool to analyze these images automatically and convert them into a chemical graph structure would be useful for many applications, such as drug discovery. A few such tools are available and they are mostly derived from optical character recognition. However, our evaluation of the performance of these tools reveals that they often make mistakes in recognizing the correct bond multiplicity and stereochemical information. In addition, errors sometimes even lead to missing atoms in the resulting graph. In our work, we address these issues by developing a compound recognition method based on machine learning. More specifically, we develop a deep neural network model for optical compound recognition. The deep learning solution presented here consists of a segmentation model, followed by three classification models that predict atom locations, bonds, and charges. Furthermore, this model not only predicts the graph structure of the molecule but also provides all information necessary to relate each component of the resulting graph to the source image. This solution is scalable and can rapidly process thousands of images. Finally, we empirically compare the proposed method with the well-established tool OSRA<sup>1</sup> and observe significant error reduction.The Supporting Information is available free of charge at <a class="ext-link" href="/doi/10.1021/acs.jcim.0c00459?goto=supporting-info">https://pubs.acs.org/doi/10.1021/acs.jcim.0c00459</a>.Eight labeled (SMILES) data sets with 1000 images each generated in four different styles, with each style, containing a set with images without stereochemical information encoded and a set with images encoded with stereochemical information (<a class="ext-link" href="/doi/suppl/10.1021/acs.jcim.0c00459/suppl_file/ci0c00459_si_001.zip">ZIP</a>)Data set with 2D images of chemical compounds cut out of journal articles, with the source article provided for every image (<a class="ext-link" href="/doi/suppl/10.1021/acs.jcim.0c00459/suppl_file/ci0c00459_si_002.zip">ZIP</a>)This article has not yet been cited by other publications.

Molecular Graph-Based Deep Learning Algorithm Facilitates an Imaging-Based Strategy for Rapid Discovery of Small Molecules Modulating Biomolecular Condensates

ComABAN: refining molecular representation with the graph attention mechanism to accelerate drug discovery

ChemGrapher: Optical Graph Recognition of Chemical Compounds by Deep Learning

Compressed graph representation for scalable molecular graph generation

Drug–target affinity prediction with extended graph learning-convolutional networks

A supervised graph-based deep learning algorithm to detect and quantify clustered particles

Graph Representation Learning for Interactive Biomolecule Systems

Molecular geometric deep learning

Learning Graph-Level Representation for Drug Discovery

A deep learning method for predicting metabolite–disease associations via graph neural network

Could graph neural networks learn better molecular representation for drug discovery? A comparison study of descriptor-based and graph-based models

Learn molecular representations from large-scale unlabeled molecules for drug discovery

Discovering Molecular Functional Groups Using Graph Convolutional Neural Networks

Molecular substructure graph attention network for molecular property identification in drug discovery

Graph neural networks for conditional de novo drug design

Advanced graph and sequence neural networks for molecular property prediction and drug discovery

Deep Learning‐Assisted Molecular Classification and Concentration Prediction by Imaging of a Large‐Area Metasurface with Spatially Gradient Geometry

An effective self-supervised framework for learning expressive molecular global representations to drug discovery

Introducing CARATE: Finally speaking chemistry through learning hidden wave-function representations on graph-based attention and convolutional neural networks

Graph networks for molecular design

A Novel Graph Neural Network Methodology to Investigate Dihydroorotate Dehydrogenase Inhibitors in Small Cell Lung Cancer