Diversity of US participants in AstraZeneca-sponsored clinical trials.

BACKGROUND: To develop medicines that are safe and efficacious to all patients, clinical trials must enroll appropriate target populations, but imbalances related to race, ethnicity and sex have been reported. A comprehensive analysis and improvement in understanding representativeness of patient enrollment in industry-sponsored trials are key public health needs. METHODS: We assessed race/ethnicity and sex representation in AstraZeneca (AZ)-sponsored clinical trials in the United States (US) from 2010 to 2022, compared with the 2019 US Census. RESULTS: In total, 246 trials representing 95,372 patients with complete race/ethnicity and sex records were analyzed. The proportions of different race/ethnicity subgroups in AZ-sponsored clinical trials and the US Census were similar (White: 69.5% vs 60.1%, Black or African American: 13.3% vs 12.5%, Asian: 1.8% vs 5.8%, Hispanic: 14.4% vs 18.5%). We also observed parity in the proportions of males and females between AZ clinical trials and US Census (males: 52.4% vs 49.2%, females: 47.6% vs 50.8%). Comparisons of four distinct therapy areas within AZ (Respiratory and Immunology [R&I]; Cardiovascular, Renal, and Metabolism [CVRM]; Solid Tumors; and Hematological Malignancies), including by trial phases, revealed greater variability, with proportions observed above and below US Census levels. CONCLUSION: This analysis provides the first detailed insights into the representativeness of AZ trials. Overall, the proportions of different race/ethnicity and sex subgroups in AZ-sponsored clinical trials were broadly aligned with the US Census. We outline some of AZ's planned health equity initiatives that are intended to continue to improve equitable patient enrollment.

Towards fully automated segmentation of rat cardiac MRI by leveraging deep learning frameworks.

Automated segmentation of human cardiac magnetic resonance datasets has been steadily improving during recent years. Similar applications would be highly useful to improve and speed up the studies of cardiac function in rodents in the preclinical context. However, the transfer of such segmentation methods to the preclinical research is compounded by the limited number of datasets and lower image resolution. In this paper we present a successful application of deep architectures 3D cardiac segmentation for rats in preclinical contexts which to our knowledge has not yet been reported. We developed segmentation models that expand on the standard U-Net architecture and evaluated models separately trained for systole and diastole phases (2MSA) and a single model trained for all phases (1MSA). Furthermore, we calibrated model outputs using a Gaussian process (GP)-based prior to improve phase selection. The resulting models approach human performance in terms of left ventricular segmentation quality and ejection fraction (EF) estimation in both 1MSA and 2MSA settings (Sørensen-Dice score 0.91 ± 0.072 and 0.93 ± 0.032, respectively). 2MSA achieved a mean absolute difference between estimated and reference EF of 3.5 ± 2.5%, while 1MSA resulted in 4.1 ± 3.0%. Applying GPs to 1MSA enabled automating systole and diastole phase selection. Both segmentation approaches (1MSA and 2MSA) were statistically equivalent. Combined with a proposed cardiac phase selection strategy, our work presents an important first step towards a fully automated segmentation pipeline in the context of rat cardiac analysis.

Siamese Recurrent Neural Network with a Self-Attention Mechanism for Bioactivity Prediction.

Activity prediction plays an essential role in drug discovery by directing search of drug candidates in the relevant chemical space. Despite being applied successfully to image recognition and semantic similarity, the Siamese neural network has rarely been explored in drug discovery where modelling faces challenges such as insufficient data and class imbalance. Here, we present a Siamese recurrent neural network model (SiameseCHEM) based on bidirectional long short-term memory architecture with a self-attention mechanism, which can automatically learn discriminative features from the SMILES representations of small molecules. Subsequently, it is used to categorize bioactivity of small molecules via N-shot learning. Trained on random SMILES strings, it proves robust across five different datasets for the task of binary or categorical classification of bioactivity. Benchmarking against two baseline machine learning models which use the chemistry-rich ECFP fingerprints as the input, the deep learning model outperforms on three datasets and achieves comparable performance on the other two. The failure of both baseline methods on SMILES strings highlights that the deep learning model may learn task-specific chemistry features encoded in SMILES strings.

Acid- and hydrogen-bonding-induced switching between 22-π and 18-π electron conjugations in 2-aminothiazolo[4,5-c]porphycenes.

2-Aminothiazolo[4,5-c]porphycenes are a novel class of 22-π electron aromatic porphycene derivatives prepared by click reaction of porphycene isothiocyanates with primary and secondary amines with high potential as near-infrared theranostic labels. Herein, the optical and photophysical properties of 2-aminothiazolo[4,5-c]porphycenes have been studied, revealing a strong dependence on hydrogen bond donor solvents and acids. High hydrogen bond donor solvents and acids shift the absorption and fluorescence emission of 2-aminothiazolo[4,5-c]porphycenes to the blue due to a contraction of their aromatic system from 22-π to 18-π electrons. Finally, the aromatic shift has been successfully used to measure the pH using 2-aminothiazoloporphycene-labelled gold nanoclusters, paving the way for the use of these compounds as near infrared pH-sensitive probes.