Research Appraisals
Evidence-based critical appraisals of the latest medical research, systematically evaluated using Oxford CEBM methodology.
Showing 10 appraisals
Journal of computer-aided molecular design
Artificial intelligence in impurity prediction: current landscape, challenges, and future directions
Earlier identification, regulation of impurities in pharmaceutical products is critical throughout the medication development process because they have a high impact on drug quality, safety, and regulatory approval. Traditional analytical methods like LC-MS and NMR are frequently used, but they are labour intensive, time consuming and have limited capacity to detect unknown or trace level impurities. New opportunities to forecast impurity formation before experimental observation have been enabled by developments in cheminformatics and computer modelling. This study covers advanced chemical representation techniques, data generation, data curation, and model validation with highlighting recent advances in machine learning and deep learning algorithms for impurity prediction. In association with forecasting synthetic by products, degradation products, and harmful contaminants, traditional QSAR techniques, and contemporary deep learning models such as graph neural networks, transformer based architectures, and generative frameworks are examined. Additionally, discussed the increasing significance of explainable models for regulatory body acceptability. Lastly, newly developed advanced techniques like digital twins, automated impurity profiling, integrated reaction degradation modelling, and real-time monitoring are examined, demonstrating how computational methods are transforming impurity assessment from a reactive effort into a predictive and preventive procedure.
23 July 2026
Read appraisal →Bioorganic chemistry
Multi-relational knowledge graph for drug-drug interaction prediction via dual aggregation and collaborative optimization
Identifying potential drug-drug interactions is crucial in clinical care and new drug development, as mutual interference between drugs can lead to adverse reactions. Recently, computational methods have been widely employed for predicting DDIs. However, these approaches often encounter high computational complexity during the process of node feature representation and frequently overlook the interactivity and polysemy of entities. Effectively handling heterogeneous information remains a significant challenge. To cope with these limitations, we propose a dual aggregation and collaborative optimization learning framework，named MRACO，for DDIs prediction based on a multi-relational knowledge graph. MRACO utilizes structural information from these graphs, efficiently learning interactivity information across various relationship types. Additionally, MRACO employs dual aggregation operations to explicitly encode and aggregate multi-type information, capturing diverse semantic relationships of drug nodes. Moreover, MRACO simplifies complex computational processes by utilizing the collaborative loss optimization function, eliminating redundant information and enhancing model robustness. By integrating deep-level interactive information, MRACO leverages rich contextual data from knowledge graphs, enhancing prediction stability. Experiments demonstrate the effective of MRACO feature extraction capabilities in multi-relational networks, highlighting its understanding of the underlying mechanisms of DDIs.
7 June 2026
Read appraisal →Psychological medicine
Risk factors associated with blood dyscrasia during clozapine treatment: a systematic review and meta-analysis
Clozapine is the only effective treatment for treatment-resistant schizophrenia, but concerns over blood dyscrasia and need for monitoring limit its use. Evidence suggests many hematological abnormalities may result from surveillance bias, with agranulocytosis being the primary adverse effect directly induced by clozapine. This systematic review (PROSPERO: CRD42024487199) investigates non-genetic risk factors associated with blood dyscrasia during clozapine treatment, focusing on neutropenia and agranulocytosis. Random-effect meta-analyses were performed on studies reporting quantitative risk data. Due to inconsistent neutropenia definitions, analyses used absolute neutrophil count (ANC) thresholds of <2000/mm3, <1500/mm3, and <500/mm3. Forty-four studies were included in the systematic review, 15 in meta-analyses. No significant association was found between agranulocytosis and female gender (OR = 1.48, 95% CI: 0.92-2.38; p = 0.106) or age (pooled standardized mean difference [SMD] = 0.32, 95% CI: -0.26 to 0.90, p = 0.285), whilst a modest inverse association with clozapine dose (SMD = -0.32, 95%CI: -0.50 to -0.14, p < 0.001) and baseline white cell count (SMD = -0.21, 95%CI: -0.40 to -0.03, p = 0.026) was found. Neutropenia (ANC < 2000/mm3) was positively associated with concomitant psychotropic use (OR = 2.15, 95% CI: 1.13-4.07, p = 0.019). Clozapine rechallenge studies revealed no significant associations with gender, age, duration of initial clozapine trial, or length of discontinuation period prior to rechallenge. No strong predictors of clozapine-associated blood dyscrasia were identified. Findings may be limited by study variability, surveillance bias, and lack of consistent differentiation between agranulocytosis and milder neutropenia, highlighting limitations in current evidence.
13 May 2026
Read appraisal →Journal of chemical information and modeling
In Silico Prediction of Potential Bile Salt Export Pump Inhibitors Using Pharmacophore Models and Consensus Machine Learning
The ATP-dependent bile salt export pump (BSEP) is a transporter responsible for moving bile salts from hepatocytes into bile canaliculi. Inhibition of BSEP is a known risk factor of cholestatic-drug-induced liver injury (DILI) probably caused by the accumulation of toxic bile salts in the liver. Since DILI is a major cause for attrition during drug development and postmarketing withdrawal or black box warnings, an IC50 of BSEP higher than 25 μM is advised. Here, we describe the investigation of several in silico approaches to map potential BSEP inhibitors. A consensus machine learning classification model is presented, which accurately predicts BSEP inhibitors and can assist in the prioritization of in vitro testing. An applicability domain was derived for the model, and it was validated by in vitro measurements of predicted BSEP inhibitors where the model was able to correctly flag eight out of 11 compounds.
13 May 2026
Read appraisal →Clinical pharmacology and therapeutics
Off-the-Shelf Large Language Models for Guiding Pharmacoepidemiological Study Design
This study aimed to assess the ability of two off-the-shelf large language models, ChatGPT and Gemini, to support the design of pharmacoepidemiological studies. We assessed 48 study protocols of pharmacoepidemiological studies published between 2018 and 2024, covering various study types, including disease epidemiology, drug utilization, safety, and effectiveness. The coherence (i.e., "Is the response coherent and well-formed, or is it difficult to understand?") and relevance (i.e., "Is the response relevant and informative, or is it lacking in substance?") of the large language models' responses were evaluated by human experts across seven key study design components. Coding accuracy was assessed. Both large language models demonstrated high coherence, with over 90% of study components rated as "Strongly agree" by experts for most categories. ChatGPT achieved the highest coherence for "Index date" (97.9%) and "Study design" (95.8%). Gemini excelled in "Study outcome" (93.9%) and "Study exposure" (95.9%). Relevance, however, was more variable, with ChatGPT aligning with expert recommendations in over 90% of cases for "Index date" and "Study design" but showing lower agreement for covariates (65%) and follow-up (70%). Coding agreement percentages reveal varying levels of concordance, with the Anatomical Therapeutic Chemical classification system coding system demonstrating the highest agreement at 50% with experts. In contrast, the Current Procedural Terminology and International Classification of Diseases systems showed agreements of 22.2% and 20%, respectively. While ChatGPT and Gemini show promise in certain tasks supporting pharmacoepidemiological study design, their limitations in relevance and coding accuracy highlight the need for critical oversight by domain experts.
4 May 2026
Read appraisal →Journal of chemical information and modeling
DWPL-GCNMF: Structure-Aware Dynamic Weighted Pseudo-Label Learning for Adverse Drug Reaction Prediction
Adverse drug reaction (ADR) prediction is typically formulated as drug-ADR association learning on extremely sparse, positive-unlabeled matrices, which limits generalization and makes naive pseudolabel expansion prone to confirmation bias. Here, we propose DWPL-GCNMF, a structure-aware semisupervised framework that combines graph convolutional network (GCN)-based structural embeddings learned from a drug-protein knowledge graph with matrix factorization scoring and introduces a prediction-consistency-driven dynamic weighted pseudolabel learning strategy to expand training pairs while down-weighting unreliable pseudolabels; multiple base models are further fused to improve prediction stability. We evaluate DWPL-GCNMF on a DrugBank-derived benchmark (1,177 drugs × 4,247 ADRs; ∼2.6% observed associations) and an external SIDER benchmark (1,080 drugs × 5,579 ADRs; ∼2.3% observed associations) to assess robustness under different data sets and increased sparsity. DWPL-GCNMF consistently outperforms representative baselines, achieving AUPR/F1 of 0.6656/0.6194 on DrugBank and 0.6106/0.5816 on SIDER, while also improving top-K prioritization (e.g., Recall@15 up to 0.8078 and 0.8146, respectively), indicating that structure-aware representation learning with dynamically weighted pseudolabeling provides a robust approach for prioritizing potential ADRs under severe data sparsity.
2 May 2026
Read appraisal →Toxicology mechanisms and methods
Introducing AutoML framework for drug-drug interaction prediction: application of AutoGluon.
Drug-drug interactions (DDIs) remain a major concern in medication safety. Although advanced artificial intelligence methods such as deep learning have improved DDI prediction, their adoption is limited by the need for specialized expertise and complex model development. This study introduces the first application of the AutoGluon AutoML framework to DDI prediction using molecular features, aiming to automate and simplify model development. A curated subset of 100,000 drug pairs from DrugBank was used, employing three molecular representations: 2D molecular descriptors, 2048-bit Morgan fingerprints, and their combination. Models were trained using AutoGluon-Tabular with no manual hyperparameter tuning. The descriptor-only model achieved the best performance, with 84.4% test accuracy and an AUC of 0.916, outperforming fingerprint-based and hybrid models. Feature importance analysis identified key physicochemical and topological descriptors-such as drug-likeness, electrotopological indices, and hydrophobic surface area-as critical predictors of DDIs. These results demonstrate that AutoML can extract chemically meaningful patterns while reducing technical barriers. Overall, our results validate AutoGluon as a scalable approach to DDI prediction that provides chemically meaningful, feature-level interpretability, and lay the groundwork for future applications involving larger datasets and more complex chemical representations.
15 Apr 2026
Read appraisal →Medicine
Real-world safety profile of anifrolumab: An observational disproportionality analysis of the FAERS database
Anifrolumab is a type I interferon receptor antagonist approved for the treatment of moderate-to-severe systemic lupus erythematosus. While its efficacy has been demonstrated in clinical trials, its long-term safety remains understudied in the real world. This study extracted relevant data from the FDA Adverse Event Reporting System (FAERS) to evaluate the post-marketing safety profile of anifrolumab in real-world clinical practice and to identify potential safety signals. We analyzed 1293 reports involving 3542 adverse events (AEs) from the FAERS database (Q3 2021-Q1 2025) using disproportionality analysis. Reports in which anifrolumab was designated as the "primary suspect" drug were included. Safety signals were identified using the reporting odds ratio and the Bayesian confidence propagation neural network, with a signal defined as positive only if it met both criteria: a lower limit of the reporting odds ratio's 95% confidence interval > 1 and the 95% lower credibility interval for the information component of the Bayesian confidence propagation neural network > 0. Additionally, multivariable logistic regression analyses were performed for key AEs to rule out potential confounding. The study identified 97 positive safety signals associated with anifrolumab. In addition to the labeled AEs, such as infusion-related reactions and infections, several unexpected AEs were also identified, including renal-related, cardiovascular-related, and neurological AEs. Multivariable logistic regression analyses for key AEs showed that concomitant immunosuppressants, comorbidities, and disease severity were not significant independent predictors (all P > .05), supporting the robustness of the identified signals. This pharmacovigilance study provides hypothesis-generating evidence on the post-marketing safety profile of anifrolumab and highlights several potential safety signals that warrant further evaluation. It is crucial to emphasize that these findings are derived from the FAERS, a spontaneous reporting database that cannot establish causality due to inherent limitations, such as underreporting, missing data, and lack of exposure information. Therefore, these results require confirmation in well-designed observational studies and clinical trials. Clinicians should be aware of these potential associations and consider appropriate monitoring.
12 Apr 2026
Read appraisal →Medicine
Post-marketing signal detection of fruquintinib-associated adverse events using the FAERS database
Fruquintinib, a targeted therapeutic agent approved in 2018 for the treatment of metastatic colorectal cancer, is gradually being applied in an expanding range of clinical settings. Given the complexities associated with real-world dosing, a comprehensive evaluation of its safety profile is essential for optimizing clinical decision-making. We conducted a retrospective pharmacovigilance study utilizing a spontaneous reporting system. Reports related to Fruquintinib were extracted through a standardized data cleaning process and cross-validated using 3 complementary signal detection algorithms: proportional reporting ratio, Bayesian confidence propagation neural network, and reporting odds ratio (ROR). This approach was employed to systematically assess the drug's safety and identify potential adverse event signals. A total of 92 statistically significant safety signals were identified from 1188 independent cases included in the analysis, corresponding to 1836 adverse event reports. The signal distribution exhibited organ system specificity, with gastrointestinal reactions being the most prevalent category. These reactions primarily manifested as typical drug-related toxicities, including diarrhea and nausea. Other notable complications included neurological, respiratory, dermatological, renal, and cardiovascular events. Important aes not mentioned in the instructions, such as bone marrow suppression (ROR = 11.17), peripheral neuropathy (ROR = 4.24) and dehydration (ROR = 3.98), were also discovered. This study systematically characterizes the post-marketing safety profile of Fruquintinib through a multi-algorithm synergistic analysis, confirming that its safety profile aligns with the clinically expected outcomes based on its mechanism of action. Future research should focus on analyzing drug interaction mechanisms and exploring biomarkers to develop a precise risk-benefit assessment model.
5 Apr 2026
Read appraisal →Heart rhythm
Machine learning-enabled assessment of risk of drug-induced QT prolongation at the time of prescribing
BACKGROUND: Many medications are associated with long QTc. Current long QTc predictors have limited generalizability and/or modest performance. OBJECTIVE: This study aimed to compare the performance of machine learning-enabled approaches to drug-induced long QTc prediction with current risk scores, including Tisdale and RISQ-PATH. METHODS: We identified patients who began QTc-prolonging medications (QTdrugs) and had follow-up 12-lead electrocardiograms (ECGs) within 1 year. Using 5-fold cross-validation, we trained XGBoost (XGB), an ECG-based deep neural network, and combined models using electronic health record data and ECG traces to predict QTc of ≥500 ms within 1 year. We assessed performance using the area under the receiver operating characteristic (AUROC) curves and compared them with corresponding Tisdale and RISQ-PATH scores. RESULTS: QTdrug records were identified for 345,371 patients, 5.7% of whom had long QTc within 1 year. In the subset with baseline ECGs available (N = 182,448; 7.7% events), both the XGB and deep neural network models demonstrated high performance (AUROC 0.869 and 0.864, respectively) but their combination yielded no significant improvement (AUROC 0.874). Therefore, focusing on the XGB model, we observed superior performance vs RISQ-PATH in the overall population (AUROC 0.859 vs 0.701), as well as Tisdale in predominantly inpatients (N = 110,558; 8.8% events; AUROC 0.855 vs 0.770). Positive predictive value was 61.5% vs 28.35% and 54.0% vs 28.3% at equivalent operating points for the XGB model vs RISQ-PATH and Tisdale, respectively. CONCLUSION: Development and retrospective validation of 3 machine learning-based models for predicting drug-induced long QTc at the time of new QTdrug starts demonstrated superior performance compared with current clinical risk calculators and may be useful tools to support medical decision making when initiating new therapies.
4 Apr 2026
Read appraisal →