Different molecular enumeration influences in deep learning: an example using aqueous solubility

INTransformer: Data augmentation-based contrastive learning by injecting noise into transformer for molecular property prediction

Molecular property prediction plays an essential role in drug discovery for identifying the candidate molecules with target properties. Deep learning models usually require sufficient labeled data to train good prediction models. However, the size of labeled data is usually small for molecular property prediction, which brings great challenges to deep learning-based molecular property prediction methods. Furthermore, the global information of molecules is critical for predicting molecular properties. Therefore, we propose INTransformer for molecular property prediction, which is a data augmentation method via contrastive learning to alleviate the limitations of the labeled molecular data while enhancing the ability to capture global information. Specifically, INTransformer consists of two identical Transformer sub-encoders to extract the molecular representation from the original SMILES and noisy SMILES respectively, while achieving the goal of data augmentation. To reduce the influence of noise, we use contrastive learning to ensure the molecular encoding of noisy SMILES is consistent with that of the original input so that the molecular representation information can be better extracted by INTransformer. Experiments on various benchmark datasets show that INTransformer achieved competitive performance for molecular property prediction tasks compared with the baselines and state-of-the-art methods.

Will we ever be able to accurately predict solubility?

Accurate prediction of thermodynamic solubility by machine learning remains a challenge. Recent models often display good performances, but their reliability may be deceiving when used prospectively. This study investigates the origins of these discrepancies, following three directions: a historical perspective, an analysis of the aqueous solubility dataverse and data quality. We investigated over 20 years of published solubility datasets and models, highlighting overlooked datasets and the overlaps between popular sets. We benchmarked recently published models on a novel curated solubility dataset and report poor performances. We also propose a workflow to cure aqueous solubility data aiming at producing useful models for bench chemist. Our results demonstrate that some state-of-the-art models are not ready for public usage because they lack a well-defined applicability domain and overlook historical data sources. We report the impact of factors influencing the utility of the models: interlaboratory standard deviation, ionic state of the solute and data sources. The herein obtained models, and quality-assessed datasets are publicly available.

Recent Studies of Artificial Intelligence on In Silico Drug Absorption

Absorption is an important area of research in pharmacochemistry and drug development, because the drug has to be absorbed before any drug effects can occur. Furthermore, the ADMET (Absorption, Distribution, Metabolism, Excretion, and Toxicity) profile of drugs can be directly and considerably altered by modulating factors affecting absorption. Many drugs in development fail because of poor absorption. The research and continuous efforts of researchers in recent years have brought many successes and promises in drug absorption property prediction, especially in silico, which helps to reduce the time and cost significantly for screening undesirable drug candidates. In this report, we explicitly provide an overview of recent in silico studies on predicting absorption properties, especially from 2019 to the present, using artificial intelligence. Additionally, we have collected and investigated public databases that support absorption prediction research. On those grounds, we also proposed the challenges and development directions of absorption prediction in the future. We hope this review can provide researchers with valuable guidelines on absorption prediction to facilitate the development of newer approaches in drug discovery.

Recent Advances in Toxicity Prediction: Applications of Deep Graph Learning

The development of new drugs is time-consuming and expensive, and as such, accurately predicting the potential toxicity of a drug candidate is crucial in ensuring its safety and efficacy. Recently, deep graph learning has become prevalent in this field due to its computational power and cost efficiency. Many novel deep graph learning methods aid toxicity prediction and further prompt drug development. This review aims to connect fundamental knowledge with burgeoning deep graph learning methods. We first summarize the essential components of deep graph learning models for toxicity prediction, including molecular descriptors, molecular representations, evaluation metrics, validation methods, and data sets. Furthermore, based on various graph-related representations of molecules, we introduce several representative studies and methods for toxicity prediction from the perspective of GNN architectures and graph pretrained models. Compared to other types of models, deep graph models not only advance in higher accuracy and efficiency but also provide more intuitive insights, which is significant in the development of model interpretation and generalization ability. The graph pretrained models are emerging as they can extract prominent features from large-scale unlabeled molecular graph data and improve the performance of downstream toxicity prediction tasks. We hope this survey can serve as a handbook for individuals interested in exploring deep graph learning for toxicity prediction.

NoiseMol: A noise-robusted data augmentation via perturbing noise for molecular property prediction

Simplified Molecular-Input Line-Entry System (SMILES) is one of a widely used molecular representation methods for molecular property prediction. We conjecture that all the characters in the SMILES string of a molecule are essential for making up the molecules, but most of them make little contribution to determining a particular property of the molecule. Therefore, we verified the conjecture in the pre-experiment. Motivated by the result, we propose to inject proper noisy information into the SMILES to augment the training data by increasing the diversity of the labeled molecules. To this end, we explore injecting perturbing noise into the original labeled SMILES strings to construct augmented data for alleviating the limitation of the labeled compound data and enhancing the model to extract more useful molecular representation for molecular property prediction. Specifically, we directly adopt mask, swap, deletion, and fusion operations on SMILES strings to randomly mask, swap, and delete atoms in SMILES strings. Then, the augmented data is used by two strategies: each epoch alternately feeds the original and perturbing noisy molecules, or each batch alternately feeds the original and perturbing noisy molecules. We conduct experiments on both Transformer and BiGRU models to validate the effectiveness by adopting widely used datasets from MoleculeNet and ZINC. Experimental results demonstrate that the proposed method outperforms strong baselines on all the datasets. NoiseMol obtains the best performance on BBBP and FDA when compared with state-of-the-art methods. Besides, NoiseMol achieves the best accuracy on LogP. Therefore, injecting perturbing noise into the labeled SMILES strings is an effective and efficient method, which improves the prediction performance, generalization, and robustness of the deep learning models.

When Machine Learning and Deep Learning Come to the Big Data in Food Chemistry

Since the first food database was released over one hundred years ago, food databases have become more diversified, including food composition databases, food flavor databases, and food chemical compound databases. These databases provide detailed information about the nutritional compositions, flavor molecules, and chemical properties of various food compounds. As artificial intelligence (AI) is becoming popular in every field, AI methods can also be applied to food industry research and molecular chemistry. Machine learning and deep learning are valuable tools for analyzing big data sources such as food databases. Studies investigating food compositions, flavors, and chemical compounds with AI concepts and learning methods have emerged in the past few years. This review illustrates several well-known food databases, focusing on their primary contents, interfaces, and other essential features. We also introduce some of the most common machine learning and deep learning methods. Furthermore, a few studies related to food databases are given as examples, demonstrating their applications in food pairing, food-drug interactions, and molecular modeling. Based on the results of these applications, it is expected that the combination of food databases and AI will play an essential role in food science and food chemistry.

Intelligent De Novo Design of Novel Antimicrobial Peptides against Antibiotic-Resistant Bacteria Strains

Because of the growing number of clinical antibiotic resistance cases in recent years, novel antimicrobial peptides (AMPs) may be ideal for next-generation antibiotics. This study trained a Wasserstein generative adversarial network with gradient penalty (WGAN-GP) based on known AMPs to generate novel AMP candidates. The quality of the GAN-designed peptides was evaluated in silico, and eight of them, named GAN-pep 1–8, were selected by an AMP Artificial Intelligence (AI) classifier and synthesized for further experiments. Disc diffusion testing and minimum inhibitory concentration (MIC) determinations were used to identify the antibacterial effects of the synthesized GAN-designed peptides. Seven of the eight synthesized GAN-designed peptides displayed antibacterial activity. Additionally, GAN-pep 3 and GAN-pep 8 presented a broad spectrum of antibacterial effects and were effective against antibiotic-resistant bacteria strains, such as methicillin-resistant Staphylococcus aureus and carbapenem-resistant Pseudomonas aeruginosa. GAN-pep 3, the most promising GAN-designed peptide candidate, had low MICs against all the tested bacteria. In brief, our approach shows an efficient way to discover AMPs effective against general and antibiotic-resistant bacteria strains. In addition, such a strategy also allows other novel functional peptides to be quickly designed, identified, and synthesized for validation on the wet bench.

On Approximating the pIC50 Value of COVID-19 Medicines In Silico with Artificial Neural Networks

In the case of pandemics such as COVID-19, the rapid development of medicines addressing the symptoms is necessary to alleviate the pressure on the medical system. One of the key steps in medicine evaluation is the determination of pIC50 factor, which is a negative logarithmic expression of the half maximal inhibitory concentration (IC50). Determining this value can be a lengthy and complicated process. A tool allowing for a quick approximation of pIC50 based on the molecular makeup of medicine could be valuable. In this paper, the creation of the artificial intelligence (AI)-based model is performed using a publicly available dataset of molecules and their pIC50 values. The modeling algorithms used are artificial and convolutional neural networks (ANN and CNN). Three approaches are tested-modeling using just molecular properties (MP), encoded SMILES representation of the molecule, and the combination of both input types. Models are evaluated using the coefficient of determination (R2) and mean absolute percentage error (MAPE) in a five-fold cross-validation scheme to assure the validity of the results. The obtained models show that the highest quality regression (R2¯=0.99, σR2¯=0.001; MAPE¯=0.009%, σMAPE¯=0.009), by a large margin, is obtained when using a hybrid neural network trained with both MP and SMILES.

Deep learning methods for molecular representation and property prediction

With advances in artificial intelligence (AI) methods, computer-aided drug design (CADD) has developed rapidly in recent years. Effective molecular representation and accurate property prediction are crucial tasks in CADD workflows. In this review, we summarize contemporary applications of deep learning (DL) methods for molecular representation and property prediction. We categorize DL methods according to the format of molecular data (1D, 2D, and 3D). In addition, we discuss some common DL models, such as ensemble learning and transfer learning, and analyze the interpretability methods for these models. We also highlight the challenges and opportunities of DL methods for molecular representation and property prediction.

Analysis of Training and Seed Bias in Small Molecules Generated with a Conditional Graph-Based Variational Autoencoder─Insights for Practical AI-Driven Molecule Generation

The application of deep learning to generative molecule design has shown early promise for accelerating lead series development. However, questions remain concerning how factors like training, data set, and seed bias impact the technology's utility to medicinal and computational chemists. In this work, we analyze the impact of seed and training bias on the output of an activity-conditioned graph-based variational autoencoder (VAE). Leveraging a massive, labeled data set corresponding to the dopamine D2 receptor, our graph-based generative model is shown to excel in producing desired conditioned activities and favorable unconditioned physical properties in generated molecules. We implement an activity-swapping method that allows for the activation, deactivation, or retention of activity of molecular seeds, and we apply independent deep learning classifiers to verify the generative results. Overall, we uncover relationships between noise, molecular seeds, and training set selection across a range of latent-space sampling procedures, providing important insights for practical AI-driven molecule generation.

A general optimization protocol for molecular property prediction using a deep learning network

The key to generating the best deep learning model for predicting molecular property is to test and apply various optimization methods. While individual optimization methods from different past works outside the pharmaceutical domain each succeeded in improving the model performance, better improvement may be achieved when specific combinations of these methods and practices are applied. In this work, three high-performance optimization methods in the literature that have been shown to dramatically improve model performance from other fields are used and discussed, eventually resulting in a general procedure for generating optimized CNN models on different properties of molecules. The three techniques are the dynamic batch size strategy for different enumeration ratios of the SMILES representation of compounds, Bayesian optimization for selecting the hyperparameters of a model and feature learning using chemical features obtained by a feedforward neural network, which are concatenated with the learned molecular feature vector. A total of seven different molecular properties (water solubility, lipophilicity, hydration energy, electronic properties, blood-brain barrier permeability and inhibition) are used. We demonstrate how each of the three techniques can affect the model and how the best model can generally benefit from using Bayesian optimization combined with dynamic batch size tuning.

Machine learning and deep learning enabled fuel sooting tendency prediction from molecular structure

Soot formation models become increasingly important in advanced renewable fuels formulation for soot reduction benefit. This work evaluates performance of machine learning (ML) and deep learning (DL) to predict yield sooting index (YSI) from chemical structure and proposes a tailor-made convolution neural network (CNN)-SDSeries38 for regression problem. In ML, a novel quantitative structure-property relationship (QSPR) is developed for feature extraction and the relationship between molecular structure and YSI is built by ML algorithm. In DL, SDSeries38 contains 9 feature learning modules, 1 regression module for automated feature learning and regression. It adopts standard series network architecture and modular structure, each feature learning module is a stack of convolution, batch normalization, activation, pooling layers. ML-QSPR model outperforms SDSeries38 in accuracy (RMSE = 7.563 vs 19.58), computational speed and the former applies to fuel mixtures. In DL, SDSeries38 network exceeds 10 classical CNN and provides a generic architecture enabling transfer application to other regression problem. DL application to regression is still in its infancy and there is no complete guide on how to develop specific CNN architectures for regression. Some gaps need to be filled: (1) Specially developed CNN architectures for regression are required; (2) The performances of direct transfer learning the classical CNN architectures from classification to regression are modest. A modular structure with typical function modules may provide an ideal solution; (3) Going deeper into the sequence of convolution layers improves predictive accuracy, but bears in mind to keep the number of layers below the threshold to avoid vanishing gradient.

Detection of microalgae objects based on the Improved YOLOv3 model

Microalgae play a major role in the invasion of alien organisms with ballast water as a carrier, and traditional ballast water detection methods have many limitations in identifying microalgae species. Therefore, this paper proposes a method to identify microalgae in ballast water based on an Improved YOLOv3 model. The method first used a lightweight network MobileNet instead of the Darknet-53 network as the backbone network of feature extraction in the original YOLOv3 model. Secondly, improved spatial pyramid pooling (SPP) is introduced to pool and concatenate the multi-scale regional features so as to reduce the position error when detecting small objects. Then, by considering the overlap area of the bounding box, central point distance and aspect ratio, the Complete IoU (CIoU) algorithm is used to optimize the loss function of the YOLOv3 model. Finally, the proposed method is experimentally compared with other latest methods on the established dataset. The experimental results demonstrated that under the same conditions, this Improved YOLOv3 model achieves an average accuracy of 98.90%, and the detection efficiency is 8.59% higher than that of the original YOLOv3 model and is better than the existing methods. The average time of this method to identify a single image is 0.086 s, and it has a good detection effect on the identification of microalgae species.

Convolutional neural networks (CNNs): concepts and applications in pharmacogenomics

Convolutional neural networks (CNNs) have been used to extract information from various datasets of different dimensions. This approach has led to accurate interpretations in several subfields of biological research, like pharmacogenomics, addressing issues previously faced by other computational methods. With the rising attention for personalized and precision medicine, scientists and clinicians have now turned to artificial intelligence systems to provide them with solutions for therapeutics development. CNNs have already provided valuable insights into biological data transformation. Due to the rise of interest in precision and personalized medicine, in this review, we have provided a brief overview of the possibilities of implementing CNNs as an effective tool for analyzing one-dimensional biological data, such as nucleotide and protein sequences, as well as small molecular data, e.g., simplified molecular-input line-entry specification, InChI, binary fingerprints, etc., to categorize the models based on their objective and also highlight various challenges. The review is organized into specific research domains that participate in pharmacogenomics for a more comprehensive understanding. Furthermore, the future intentions of deep learning are outlined.

Improvement of Prediction Performance With Conjoint Molecular Fingerprint in Deep Learning

The accurate predicting of physical properties and bioactivity of drug molecules in deep learning depends on how molecules are represented. Many types of molecular descriptors have been developed for quantitative structure-activity/property relationships quantitative structure-activity relationships (QSPR). However, each molecular descriptor is optimized for a specific application with encoding preference. Considering that standalone featurization methods may only cover parts of information of the chemical molecules, we proposed to build the conjoint fingerprint by combining two supplementary fingerprints. The impact of conjoint fingerprint and each standalone fingerprint on predicting performance was systematically evaluated in predicting the logarithm of the partition coefficient (logP) and binding affinity of protein-ligand by using machine learning/deep learning (ML/DL) methods, including random forest (RF), support vector regression (SVR), extreme gradient boosting (XGBoost), long short-term memory network (LSTM), and deep neural network (DNN). The results demonstrated that the conjoint fingerprint yielded improved predictive performance, even outperforming the consensus model using two standalone fingerprints among four out of five examined methods. Given that the conjoint fingerprint scheme shows easy extensibility and high applicability, we expect that the proposed conjoint scheme would create new opportunities for continuously improving predictive performance of deep learning by harnessing the complementarity of various types of fingerprints.

MRFF-YOLO: A Multi-Receptive Fields Fusion Network for Remote Sensing Target Detection

High-altitude remote sensing target detection has problems related to its low precision and low detection rate. In order to enhance the performance of detecting remote sensing targets, a new YOLO (You Only Look Once)-V3-based algorithm was proposed. In our improved YOLO-V3, we introduced the concept of multi-receptive fields to enhance the performance of feature extraction. Therefore, the proposed model was termed Multi-Receptive Fields Fusion YOLO (MRFF-YOLO). In addition, to address the flaws of YOLO-V3 in detecting small targets, we increased the detection layers from three to four. Moreover, in order to avoid gradient fading, the structure of improved DenseNet was chosen in the detection layers. We compared our approach (MRFF-YOLO) with YOLO-V3 and other state-of-the-art target detection algorithms on an Remote Sensing Object Detection (RSOD) dataset and a dataset of Object Detection in Aerial Images (UCS-AOD). With a series of improvements, the mAP (mean average precision) of MRFF-YOLO increased from 77.10% to 88.33% in the RSOD dataset and increased from 75.67% to 90.76% in the UCS-AOD dataset. The leaking detection rates are also greatly reduced, especially for small targets. The experimental results showed that our approach achieved better performance than traditional YOLO-V3 and other state-of-the-art models for remote sensing target detection.