Prediction Of Adverse Drug Events Research Articles

MtADENet: A novel interpretable method integrating multiple types of network-based inference approaches for prediction of adverse drug events

Identification of adverse drug events (ADEs) is crucial to reduce human health risks and accelerate drug safety assessment. ADEs are mainly caused by unintended interactions with primary or additional targets (off-targets). In this study, we proposed a novel interpretable method named mtADENet, which integrates multiple types of network-based inference approaches for ADE prediction. Different from phenotype-based methods, mtADENet introduced computational target profiles predicted by network-based methods to bridge the gap between chemical structures and ADEs, and hence can not only predict ADEs for drugs and novel compounds within or outside the drug-ADE association network, but also provide insights for the elucidation of molecular mechanisms of the ADEs caused by drugs. We constructed a series of network-based prediction models for 23 ADE categories. These models achieved high AUC values ranging from 0.865 to 0.942 in 10-fold cross validation. The best model further showed high performance on four external validation sets, which outperformed two previous network-based methods. To show the practical value of mtADENet, we performed case studies on developmental neurotoxicity and cardio-oncology, and over 50 % of predicted ADEs and targets for drugs and novel compounds were validated by literature. Moreover, mtADENet is freely available at our web server named NetInfer (http://lmmd.ecust.edu.cn/netinfer/). In summary, mtADENet would be a powerful tool for ADE prediction and drug safety assessment in drug discovery and development.

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.

Neural Collective Matrix Factorization for integrated analysis of heterogeneous biomedical data.

In many biomedical studies, there arises the need to integrate data from multiple directly or indirectly related sources. Collective matrix factorization (CMF) and its variants are models designed to collectively learn from arbitrary collections of matrices. The latent factors learnt are rich integrative representations that can be used in downstream tasks, such as clustering or relation prediction with standard machine-learning models. Previous CMF-based methods have numerous modeling limitations. They do not adequately capture complex non-linear interactions and do not explicitly model varying sparsity and noise levels in the inputs, and some cannot model inputs with multiple datatypes. These inadequacies limit their use on many biomedical datasets. To address these limitations, we develop Neural Collective Matrix Factorization (NCMF), the first fully neural approach to CMF. We evaluate NCMF on relation prediction tasks of gene-disease association prediction and adverse drug event prediction, using multiple datasets. In each case, data are obtained from heterogeneous publicly available databases and used to learn representations to build predictive models. NCMF is found to outperform previous CMF-based methods and several state-of-the-art graph embedding methods for representation learning in our experiments. Our experiments illustrate the versatility and efficacy of NCMF in representation learning for seamless integration of heterogeneous data. https://github.com/ajayago/NCMF_bioinformatics. Supplementary data are available at Bioinformatics online.

Adverse Drug Event Prediction Using Noisy Literature-Derived Knowledge Graphs: Algorithm Development and Validation.

BackgroundAdverse drug events (ADEs) are unintended side effects of drugs that cause substantial clinical and economic burdens globally. Not all ADEs are discovered during clinical trials; therefore, postmarketing surveillance, called pharmacovigilance, is routinely conducted to find unknown ADEs. A wealth of information, which facilitates ADE discovery, lies in the growing body of biomedical literature. Knowledge graphs (KGs) encode information from the literature, where the vertices and the edges represent clinical concepts and their relations, respectively. The scale and unstructured form of the literature necessitates the use of natural language processing (NLP) to automatically create such KGs. Previous studies have demonstrated the utility of such literature-derived KGs in ADE prediction. Through unsupervised learning of the representations (features) of clinical concepts from the KG, which are used in machine learning models, state-of-the-art results for ADE prediction were obtained on benchmark data sets.ObjectiveDue to the use of NLP to infer literature-derived KGs, there is noise in the form of false positive (erroneous) and false negative (absent) nodes and edges. Previous representation learning methods do not account for such inaccuracies in the graph. NLP algorithms can quantify the confidence in their inference of extracted concepts and relations from the literature. Our hypothesis, which motivates this work, is that by using such confidence scores during representation learning, the learned embeddings would yield better features for ADE prediction models.MethodsWe developed methods to use these confidence scores on two well-known representation learning methods—DeepWalk and Translating Embeddings for Modeling Multi-relational Data (TransE)—to develop their weighted versions: Weighted DeepWalk and Weighted TransE. These methods were used to learn representations from a large literature-derived KG, the Semantic MEDLINE Database, which contains more than 93 million clinical relations. They were compared with Embedding of Semantic Predications, which, to our knowledge, is the best reported representation learning method using the Semantic MEDLINE Database with state-of-the-art results for ADE prediction. Representations learned from different methods were used (separately) as features of drugs and diseases to build classification models for ADE prediction using benchmark data sets. The methods were compared rigorously over multiple cross-validation settings.ResultsThe weighted versions we designed were able to learn representations that yielded more accurate predictive models than the corresponding unweighted versions of both DeepWalk and TransE, as well as Embedding of Semantic Predications, in our experiments. There were performance improvements of up to 5.75% in the F1-score and 8.4% in the area under the receiver operating characteristic curve value, thus advancing the state of the art in ADE prediction from literature-derived KGs.ConclusionsOur classification models can be used to aid pharmacovigilance teams in detecting potentially new ADEs. Our experiments demonstrate the importance of modeling inaccuracies in the inferred KGs for representation learning.

Improved Adverse Drug Event Prediction Through Information Component Guided Pharmacological Network Model (IC-PNM).

Improving adverse drug event (ADE) prediction is highly critical in pharmacovigilance research. We propose a novel information component guided pharmacological network model (IC-PNM) to predict drug-ADE signals. This new method combines the pharmacological network model and information component, a Bayes statistics method. We use 33,947 drug-ADE pairs from the FDA Adverse Event Reporting System (FAERS) 2010 data as the training data, and the new 21,065 drug-ADE pairs from FAERS 2011-2015 as the validations samples. The IC-PNM data analysis suggests that both large and small sample size drug-ADE pairs are needed in training the predictive model for its prediction performance to reach an area under the receiver operating characteristic curve [Formula: see text]. On the other hand, the IC-PNM prediction performance improved to [Formula: see text] if we removed the small sample size drug-ADE pairs from the prediction model during validation.

Exploiting complex medical data with interpretable deep learning for adverse drug event prediction.

A variety of deep learning architectures have been developed for the goal of predictive modelling and knowledge extraction from medical records. Several models have placed strong emphasis on temporal attention mechanisms and decay factors as a means to include highly temporally relevant information regarding the recency of medical event occurrence while facilitating medical code-level interpretability. In this study we utilise such models with a large Electronic Patient Record (EPR) data set consisting of diagnoses, medication, and clinical text data for the purpose of adverse drug event (ADE) prediction. The first contribution of this work is an empirical evaluation of two state-of-the-art medical-code based models in terms of objective performance metrics for ADE prediction on diagnosis and medication data. Secondly, as an extension of previous work, we augment an interpretable deep learning architecture to permit numerical risk and clinical text features and demonstrate how this approach yields improved predictive performance compared to the other baselines. Finally, we assess the importance of attention mechanisms in regards to their usefulness for medical code-level and text-level interpretability, which may facilitate novel insights pertaining to the nature of ADE occurrence within the health care domain.

Improving the Prediction of Adverse Drug Events Using Feature Fusion-Based Predictive Network Models

Computational strategies play a vital role in the prediction of adverse drug events (ADEs) owing to their low cost and increased efficiency. In this study, we used the strengths of the Jaccard and Adamic-Adar indices to build feature fusion-based predictive network models (FFPNMs) with three different machine learning (ML) methods respectively to predict drug-ADE associations. Our FFPNM with the logistic regression (LR) model improved to an area under the receiver operating characteristic curve (AUROC) value of 0.849, while the corresponding AUROC values for the pharmacological network model (PNM) and model based on similarity measures were 0.824 and 0.821, respectively. FFPNM with random forest (RF) is the best model among them with an AUROC value of 0.856, and the performance of FFPNM with SVM is close to that of FFPNM with RF and higher than that of FFPNM with LR. In these models, the bipartite network consisted of 152 drugs and 633 ADEs, which were obtained from the FDA Adverse Event Reporting System (FAERS) 2010 dataset. To better evaluate the performance of FFPNMs, we performed model predictions by different network consisting of 1177 drugs and 97 ADEs which were from the data of the first 120 days of FAERS 2004. FFPNM with RF achieved the best predictive result with AUROC value of 0.913. The results show that FFPNMs with ML methods, specially RF, have a superior prediction performance and robustness using only the topology features of the drug-ADE network. From our findings, the optimal, concise, and efficient models as computational methods for drug-ADE association predictions, were revealed. Source codes of this paper are available on https://github.com/Coderljl/FFPNM.

Mendelian randomization: a novel approach for the prediction of adverse drug events and drug repurposing opportunities.

Identification of unintended drug effects, specifically drug repurposing opportunities and adverse drug events, maximizes the benefit of a drug and protects the health of patients. However, current observational research methods are subject to several biases. These include confounding by indication, reverse causality and missing data. We propose that Mendelian randomization (MR) offers a novel approach for the prediction of unintended drug effects. In particular, we advocate the synthesis of evidence from this method and other approaches, in the spirit of triangulation, to improve causal inferences concerning drug effects. MR addresses some of the limitations associated with the existing methods in this field. Furthermore, it can be applied either before or after approval of the drug, and could therefore prevent the potentially harmful exposure of patients in clinical trials and beyond. The potential of MR as a pharmacovigilance and drug repurposing tool is yet to be realized, and could both help prevent adverse drug events and identify novel indications for existing drugs in the future.

Concordance and predictive value of two adverse drug event data sets.

BackgroundAccurate prediction of adverse drug events (ADEs) is an important means of controlling and reducing drug-related morbidity and mortality. Since no single “gold standard” ADE data set exists, a range of different drug safety data sets are currently used for developing ADE prediction models. There is a critical need to assess the degree of concordance between these various ADE data sets and to validate ADE prediction models against multiple reference standards.MethodsWe systematically evaluated the concordance of two widely used ADE data sets – Lexi-comp from 2010 and SIDER from 2012. The strength of the association between ADE (drug) counts in Lexi-comp and SIDER was assessed using Spearman rank correlation, while the differences between the two data sets were characterized in terms of drug categories, ADE categories and ADE frequencies. We also performed a comparative validation of the Predictive Pharmacosafety Networks (PPN) model using both ADE data sets. The predictive power of PPN using each of the two validation sets was assessed using the area under Receiver Operating Characteristic curve (AUROC).ResultsThe correlations between the counts of ADEs and drugs in the two data sets were 0.84 (95% CI: 0.82-0.86) and 0.92 (95% CI: 0.91-0.93), respectively. Relative to an earlier snapshot of Lexi-comp from 2005, Lexi-comp 2010 and SIDER 2012 introduced a mean of 1,973 and 4,810 new drug-ADE associations per year, respectively. The difference between these two data sets was most pronounced for Nervous System and Anti-infective drugs, Gastrointestinal and Nervous System ADEs, and postmarketing ADEs. A minor difference of 1.1% was found in the AUROC of PPN when SIDER 2012 was used for validation instead of Lexi-comp 2010.ConclusionsIn conclusion, the ADE and drug counts in Lexi-comp and SIDER data sets were highly correlated and the choice of validation set did not greatly affect the overall prediction performance of PPN. Our results also suggest that it is important to be aware of the differences that exist among ADE data sets, especially in modeling applications focused on specific drug and ADE categories.

Current understanding of hepatic and intestinal OATP-mediated drug–drug interactions

At present, many patients are medicated with various drugs, which are, at the same time, associated with an increased risk of drug–drug interactions (DDIs). Detailed analysis of mechanisms underlying DDIs is the basis of a better prediction of adverse drug events caused by drug interactions. In the last few decades, an involvement of transporters in such processes has been more and more recognized. Indeed, uptake transporters belonging to the organic anion-transporting polypeptide (OATP) family have been shown to interact with a variety of drugs in clinical use. Particularly, the subfamily of OATP1B transporters has been extensively studied, identifying several clinical significant DDIs based on those hepatic uptake transporters. By contrast, the role of OATP2B1 in this context is rather underestimated. Therefore, in addition to known interactions based on OATP1B transporters, we have focused on DDIs probably based on OATP2B1 inhibition in the liver and those possibly owing to the inhibition of OATP2B1-mediated drug absorption in the intestine.

Integration of Diverse Data Sources for Prediction of Adverse Drug Events

The rapid evolution of large biological, pharmacological, and chemical databases has led to optimism that such data resources can be leveraged for prediction of drug action based on molecular descriptors of the drug. Challenges to realize this possibility include organization of each type of database in a manner that allows extraction of information across disparate data sources and the linkage of information across the biological, pharmacological, and chemical domains.

A strategy for the detection of quality fluctuation of a Chinese herbal injection based on chemical fingerprinting combined with biological fingerprinting

Herbal injections are powerful preparations of traditional Chinese medicines. However, the quality control (QC) of these formulations is difficult to establish. Recently, chemical fingerprinting (CF) has been recommended as a potential strategy for the QC of herbal injections. However, some constituents cannot be detected by chromatographic methods. To establish a comprehensive QC process, biological fingerprinting (BF) was combined with CF to detect the fluctuation in quality of a herbal injection from chemical and biological aspects. Yinzhihuang injection was selected as a representative herbal injection. Ten batches of normal samples and six batches of artificially abnormal samples were collected. High-performance liquid chromatography and thermal activity monitoring were applied to develop CF and BF, respectively. The CF and BF of normal samples were similar, with good stability and consistency, but the abnormal samples were not. The results were analyzed by hierarchical clustering analysis: all abnormal samples could be correctly distinguished when CF and BF were combined. The recognition ratio was higher for CF (87.5%) than for BF (93.75%). The key factors of quality fluctuation were bacterial contamination, high temperature, lighting, and sterilizing conditions. The described method could be used for early prediction of adverse drug events, which could help improve the safety of herbal injections.