Quantitative prediction of drug side effects based on drug-related features

Crossfeat: a transformer-based cross-feature learning model for predicting drug side effect frequency

BACKGROUND

Safe drug treatment requires an understanding of the potential side effects. Identifying the frequency of drug side effects can reduce the risks associated with drug use. However, existing computational methods for predicting drug side effect frequencies heavily depend on known drug side effect frequency information. Consequently, these methods face challenges when predicting the side effect frequencies of new drugs. Although a few methods can predict the side effect frequencies of new drugs, they exhibit unreliable performance owing to the exclusion of drug-side effect relationships.

RESULTS

This study proposed CrossFeat, a model based on convolutional neural network-transformer architecture with cross-feature learning that can predict the occurrence and frequency of drug side effects for new drugs, even in the absence of information regarding drug-side effect relationships. CrossFeat facilitates the concurrent learning of drugs and side effect information within its transformer architecture. This simultaneous exchange of information enables drugs to learn about their associated side effects, while side effects concurrently acquire information about the respective drugs. Such bidirectional learning allows for the comprehensive integration of drug and side effect knowledge. Our five-fold cross-validation experiments demonstrated that CrossFeat outperforms existing studies in predicting side effect frequencies for new drugs without prior knowledge.

CONCLUSIONS

Our model offers a promising approach for predicting the drug side effect frequencies, particularly for new drugs where prior information is limited. CrossFeat's superior performance in cross-validation experiments, along with evidence from case studies and ablation experiments, highlights its effectiveness.

Machine Learning Techniques for Predicting Drug-Related Side Effects: A Scoping Review

BACKGROUND

Drug safety relies on advanced methods for timely and accurate prediction of side effects. To tackle this requirement, this scoping review examines machine-learning approaches for predicting drug-related side effects with a particular focus on chemical, biological, and phenotypical features.

METHODS

This was a scoping review in which a comprehensive search was conducted in various databases from 1 January 2013 to 31 December 2023.

RESULTS

The results showed the widespread use of Random Forest, k-nearest neighbor, and support vector machine algorithms. Ensemble methods, particularly random forest, emphasized the significance of integrating chemical and biological features in predicting drug-related side effects.

CONCLUSIONS

This review article emphasized the significance of considering a variety of features, datasets, and machine learning algorithms for predicting drug-related side effects. Ensemble methods and Random Forest showed the best performance and combining chemical and biological features improved prediction. The results suggested that machine learning techniques have some potential to improve drug development and trials. Future work should focus on specific feature types, selection techniques, and graph-based methods for even better prediction.

MultiGML: Multimodal graph machine learning for prediction of adverse drug events

Adverse drug events constitute a major challenge for the success of clinical trials. Several computational strategies have been suggested to estimate the risk of adverse drug events in preclinical drug development. While these approaches have demonstrated high utility in practice, they are at the same time limited to specific information sources. Thus, many current computational approaches neglect a wealth of information which results from the integration of different data sources, such as biological protein function, gene expression, chemical compound structure, cell-based imaging and others. In this work we propose an integrative and explainable multi-modal Graph Machine Learning approach (MultiGML), which fuses knowledge graphs with multiple further data modalities to predict drug related adverse events and general drug target-phenotype associations. MultiGML demonstrates excellent prediction performance compared to alternative algorithms, including various traditional knowledge graph embedding techniques. MultiGML distinguishes itself from alternative techniques by providing in-depth explanations of model predictions, which point towards biological mechanisms associated with predictions of an adverse drug event. Hence, MultiGML could be a versatile tool to support decision making in preclinical drug development.

Identification of Chemical-Disease Associations Through Integration of Molecular Fingerprint, Gene Ontology and Pathway Information

The identification of chemical-disease association types is helpful not only to discovery lead compounds and study drug repositioning, but also to treat disease and decipher pathomechanism. It is very urgent to develop computational method for identifying potential chemical-disease association types, since wet methods are usually expensive, laborious and time-consuming. In this study, molecular fingerprint, gene ontology and pathway are utilized to characterize chemicals and diseases. A novel predictor is proposed to recognize potential chemical-disease associations at the first layer, and further distinguish whether their relationships belong to biomarker or therapeutic relations at the second layer. The prediction performance of current method is assessed using the benchmark dataset based on ten-fold cross-validation. The practical prediction accuracies of the first layer and the second layer are 78.47% and 72.07%, respectively. The recognition ability for lead compounds, new drug indications, potential and true chemical-disease association pairs has also been investigated and confirmed by constructing a variety of datasets and performing a series of experiments. It is anticipated that the current method can be considered as a powerful high-throughput virtual screening tool for drug researches and developments.

Literature based discovery of alternative TCM medicine for adverse reactions to depression drugs

Background

In recent years, Traditional Chinese Medicine (TCM) and alternative medicine have been widely used along with western drugs as a complementary form of treatment. In this study, we first use the scientific literature to identify western drugs with obvious side effects. Then, we find TCM alternatives for these western drugs to ameliorate their side effects.

Results

We used depression as a case study. To evaluate our method, we showed the relation between herb-ingredients-target-disease for representative alternative herbs of western drugs. Further, a protein-protein interaction network of western drugs and alternative herbs was produced, and we performed enrichment analysis of the targets of the active ingredients of the herbs and examined the enrichment of Gene Ontology terms for Biological Process, Cellular Component, and Molecular Function and KEGG Pathway levels, to show how these targets affect different levels of gene expression.

Conclusion

Our proposed method is able to select herbs that are highly relevant to the target indication (depression) and are able to treat the side effects caused by the target drug. The compounds from our selected alternative herbal medicines can therefore be complementary to the western drugs and ameliorate their side effects, which may help in the development of new drugs.

Prediction of Drug Side Effects with a Refined Negative Sample Selection Strategy

Drugs are an important way to treat various diseases. However, they inevitably produce side effects, bringing great risks to human bodies and pharmaceutical companies. How to predict the side effects of drugs has become one of the essential problems in drug research. Designing efficient computational methods is an alternative way. Some studies paired the drug and side effect as a sample, thereby modeling the problem as a binary classification problem. However, the selection of negative samples is a key problem in this case. In this study, a novel negative sample selection strategy was designed for accessing high-quality negative samples. Such strategy applied the random walk with restart (RWR) algorithm on a chemical-chemical interaction network to select pairs of drugs and side effects, such that drugs were less likely to have corresponding side effects, as negative samples. Through several tests with a fixed feature extraction scheme and different machine-learning algorithms, models with selected negative samples produced high performance. The best model even yielded nearly perfect performance. These models had much higher performance than those without such strategy or with another selection strategy. Furthermore, it is not necessary to consider the balance of positive and negative samples under such a strategy.

Predicting Drug Side Effects with Compact Integration of Heterogeneous Networks

Background:

The side effects of drugs are not only harmful to humans but also the major reasons for withdrawing approved drugs, bringing greater risks for pharmaceutical companies. However, detecting the side effects for a given drug via traditional experiments is time- consuming and expensive. In recent years, several computational methods have been proposed to predict the side effects of drugs. However, most of the methods cannot effectively integrate the heterogeneous properties of drugs.

Methods:

In this study, we adopted a network embedding method, Mashup, to extract essential and informative drug features from several drug heterogeneous networks, representing different properties of drugs. For side effects, a network was also built, from where side effect features were extracted. These features can capture essential information about drugs and side effects in a network level. Drug and side effect features were combined together to represent each pair of drug and side effect, which was deemed as a sample in this study. Furthermore, they were fed into a random forest (RF) algorithm to construct the prediction model, called the RF network model.

Results:

The RF network model was evaluated by several tests. The average of Matthews correlation coefficients on the balanced and unbalanced datasets was 0.640 and 0.641, respectively.

Conclusion:

The RF network model was superior to the models incorporating other machine learning algorithms and one previous model. Finally, we also investigated the influence of two feature dimension parameters on the RF network model and found that our model was not very sensitive to these parameters.

STS-NLSP: A Network-Based Label Space Partition Method for Predicting the Specificity of Membrane Transporter Substrates Using a Hybrid Feature of Structural and Semantic Similarity

Membrane transport proteins play crucial roles in the pharmacokinetics of substrate drugs, the drug resistance in cancer and are vital to the process of drug discovery, development and anti-cancer therapeutics. However, experimental methods to profile a substrate drug against a panel of transporters to determine its specificity are labor intensive and time consuming. In this article, we aim to develop an in silico multi-label classification approach to predict whether a substrate can specifically recognize one of the 13 categories of drug transporters ranging from ATP-binding cassette to solute carrier families using both structural fingerprints and chemical ontologies information of substrates. The data-driven network-based label space partition (NLSP) method was utilized to construct the model based on a hybrid of similarity-based feature by the integration of 2D fingerprint and semantic similarity. This method builds predictors for each label cluster (possibly intersecting) detected by community detection algorithms and takes union of label sets for a compound as final prediction. NLSP lies into the ensembles of multi-label classifier category in multi-label learning field. We utilized Cramér's V statistics to quantify the label correlations and depicted them via a heatmap. The jackknife tests and iterative stratification based cross-validation method were adopted on a benchmark dataset to evaluate the prediction performance of the proposed models both in multi-label and label-wise manner. Compared with other powerful multi-label methods, ML-kNN, MTSVM, and RAkELd, our multi-label classification model of NLPS-RF (random forest-based NLSP) has proven to be a feasible and effective model, and performed satisfactorily in the predictive task of transporter-substrate specificity. The idea behind NLSP method is intriguing and the power of NLSP remains to be explored for the multi-label learning problems in bioinformatics. The benchmark dataset, intermediate results and python code which can fully reproduce our experiments and results are available at https://github.com/dqwei-lab/STS.

ATC-NLSP: Prediction of the Classes of Anatomical Therapeutic Chemicals Using a Network-Based Label Space Partition Method

Anatomical Therapeutic Chemical (ATC) classification system proposed by the World Health Organization is a widely accepted drug classification scheme in both academic and industrial realm. It is a multilabeling system which categorizes drugs into multiple classes according to their therapeutic, pharmacological, and chemical attributes. In this study, we adopted a data-driven network-based label space partition (NLSP) method for prediction of ATC classes of a given compound within the multilabel learning framework. The proposed method ATC-NLSP is trained on the similarity-based features such as chemical–chemical interaction and structural and fingerprint similarities of a compound to other compounds belonging to the different ATC categories. The NLSP method trains predictors for each label cluster (possibly intersecting) detected by community detection algorithms and takes the ensemble labels for a compound as final prediction. Experimental evaluation based on the jackknife test on the benchmark dataset demonstrated that our method has boosted the absolute true rate, which is the most stringent evaluation metrics in this study, from 0.6330 to 0.7497, in comparison to the state-of-the-art approaches. Moreover, the community structures of the label relation graph were detected through the label propagation method. The advantage of multilabel learning over the single-label models was shown by label-wise analysis. Our study indicated that the proposed method ATC-NLSP, which adopts ideas from network research community and captures the correlation of labels in a data driven manner, is the top-performing model in the ATC prediction task. We believed that the power of NLSP remains to be unleashed for the multilabel learning tasks in drug discovery. The source codes are freely available at https://github.com/dqwei-lab/ATC.

Developing a Multi-Dose Computational Model for Drug-Induced Hepatotoxicity Prediction Based on Toxicogenomics Data

Drug-induced hepatotoxicity may cause acute and chronic liver disease, leading to great concern for patient safety. It is also one of the main reasons for drug withdrawal from the market. Toxicogenomics data has been widely used in hepatotoxicity prediction. In our study, we proposed a multi-dose computational model to predict the drug-induced hepatotoxicity based on gene expression and toxicity data. The dose/concentration information after drug treatment is fully utilized in our study based on the dose-response curve, thus a more informative representative of the dose-response relationship is considered. We also proposed a new feature selection method, named MEMO, which is also one important aspect of our multi-dose model in our study, to deal with the high-dimensional toxicogenomics data. We validated the proposed model using the TG-GATEs, which is a large database recording toxicogenomics data from multiple views. The experimental results show that the drug-induced hepatotoxicity can be predicted with high accuracy and efficiency using the proposed predictive model.

Survey of Machine Learning Techniques for Prediction of the Isoform Specificity of Cytochrome P450 Substrates

Background:

Determination or prediction of the Absorption, Distribution, Metabolism, and Excretion (ADME) properties of drug candidates and drug-induced toxicity plays crucial roles in drug discovery and development. Metabolism is one of the most complicated pharmacokinetic properties to be understood and predicted. However, experimental determination of the substrate binding, selectivity, sites and rates of metabolism is time- and recourse- consuming. In the phase I metabolism of foreign compounds (i.e., most of drugs), cytochrome P450 enzymes play a key role. To help develop drugs with proper ADME properties, computational models are highly desired to predict the ADME properties of drug candidates, particularly for drugs binding to cytochrome P450.

Objective:

This narrative review aims to briefly summarize machine learning techniques used in the prediction of the cytochrome P450 isoform specificity of drug candidates.

Results:

Both single-label and multi-label classification methods have demonstrated good performance on modelling and prediction of the isoform specificity of substrates based on their quantitative descriptors.

Conclusion:

This review provides a guide for researchers to develop machine learning-based methods to predict the cytochrome P450 isoform specificity of drug candidates.

A similarity-based method for prediction of drug side effects with heterogeneous information

Drugs can produce intended therapeutic effects to treat different diseases. However, they may also cause side effects at the same time. For an approved drug, it is best to detect all side effects it can produce. Otherwise, it may bring great risks for pharmaceuticals companies as well as be harmful to human body. It is urgent to design quick and reliable identification methods to detect the side effects for a given drug. In this study, a binary classification model was proposed to predict drug side effects. Different from most previous methods, our model termed the pair of drug and side effect as a sample and convert the original problem to a binary classification problem. Based on the similarity idea, each pair was represented by five features, each of which was derived from a type of drug property. The strong machine learning algorithm, random forest, was adopted as the prediction engine. The ten-fold cross-validation on five datasets with different negative samples indicated that the proposed model yielded a good performance of Matthews correlation coefficient around 0.550 and AUC around 0.8492. In addition, we also analyzed the contribution of each drug property for construction of the model. The results indicated that drug similarity in fingerprint was most related to the prediction of drug side effects and all drug properties gave less or more contributions.

Prediction of Effective Drug Combinations by an Improved Naïve Bayesian Algorithm

Drug combinatorial therapy is a promising strategy for combating complex diseases due to its fewer side effects, lower toxicity and better efficacy. However, it is not feasible to determine all the effective drug combinations in the vast space of possible combinations given the increasing number of approved drugs in the market, since the experimental methods for identification of effective drug combinations are both labor- and time-consuming. In this study, we conducted systematic analysis of various types of features to characterize pairs of drugs. These features included information about the targets of the drugs, the pathway in which the target protein of a drug was involved in, side effects of drugs, metabolic enzymes of the drugs, and drug transporters. The latter two features (metabolic enzymes and drug transporters) were related to the metabolism and transportation properties of drugs, which were not analyzed or used in previous studies. Then, we devised a novel improved naïve Bayesian algorithm to construct classification models to predict effective drug combinations by using the individual types of features mentioned above. Our results indicated that the performance of our proposed method was indeed better than the naïve Bayesian algorithm and other conventional classification algorithms such as support vector machine and K-nearest neighbor.