Explore the vast range of titles available.

RNAi therapeutics have been growing. Patisiran and givosiran, two siRNA-based drugs, were approved by the Food and Drug Administration in 2018 and 2019, respectively. However, there is rare news on the advance of miRNA drugs (another therapeutic similar to siRNA drug). Here we report the existing obstacles of miRNA therapeutics by analyses for resources available in a drug target perspective, despite being appreciated when it began. Only 10 obtainable miRNA drugs have been in clinical trials with none undergoing phase III, while over 60 siRNA drugs are in complete clinical trial progression including two approvals. We mechanically compared the two types of drug and found that their major distinction lay in the huge discrepancy of the target number of two RNA molecules, which was caused by different complementary ratios. One miRNA generally targets tens and even hundreds of genes. We named it \"too many targets for miRNA effect\" (TMTME). Further, two adverse events from the discontinuation of two miRNA therapeutics were exactly answered by TMTME. In summary, TMTME is inevitable because of the special complementary approach between miRNA and its target. It means that miRNA therapeutics would trigger a series of unknown and unpreventable consequences, which makes it a considerable alternative for application.

A knowledge gap persists between machine learning (ML) developers (e.g., data scientists) and practitioners (e.g., clinicians), hampering the full utilization of ML for clinical data analysis. We investigated the potential of the ChatGPT Advanced Data Analysis (ADA), an extension of GPT-4, to bridge this gap and perform ML analyses efficiently. Real-world clinical datasets and study details from large trials across various medical specialties were presented to ChatGPT ADA without specific guidance. ChatGPT ADA autonomously developed state-of-the-art ML models based on the original study’s training data to predict clinical outcomes such as cancer development, cancer progression, disease complications, or biomarkers such as pathogenic gene sequences. Following the re-implementation and optimization of the published models, the head-to-head comparison of the ChatGPT ADA-crafted ML models and their respective manually crafted counterparts revealed no significant differences in traditional performance metrics ( p  ≥ 0.072). Strikingly, the ChatGPT ADA-crafted ML models often outperformed their counterparts. In conclusion, ChatGPT ADA offers a promising avenue to democratize ML in medicine by simplifying complex data analyses, yet should enhance, not replace, specialized training and resources, to promote broader applications in medical research and practice. A knowledge gap persists between machine learning developers and clinicians. Here, the authors show that the Advanced Data Analysis extension of ChatGPT could bridge this gap and simplify complex data analyses, making them more accessible to clinicians.

Although generative adversarial networks (GANs) can produce large datasets, their limited diversity and fidelity have been recently addressed by denoising diffusion probabilistic models, which have demonstrated superiority in natural image synthesis. In this study, we introduce Medfusion, a conditional latent DDPM designed for medical image generation, and evaluate its performance against GANs, which currently represent the state-of-the-art. Medfusion was trained and compared with StyleGAN-3 using fundoscopy images from the AIROGS dataset, radiographs from the CheXpert dataset, and histopathology images from the CRCDX dataset. Based on previous studies, Progressively Growing GAN (ProGAN) and Conditional GAN (cGAN) were used as additional baselines on the CheXpert and CRCDX datasets, respectively. Medfusion exceeded GANs in terms of diversity (recall), achieving better scores of 0.40 compared to 0.19 in the AIROGS dataset, 0.41 compared to 0.02 (cGAN) and 0.24 (StyleGAN-3) in the CRMDX dataset, and 0.32 compared to 0.17 (ProGAN) and 0.08 (StyleGAN-3) in the CheXpert dataset. Furthermore, Medfusion exhibited equal or higher fidelity (precision) across all three datasets. Our study shows that Medfusion constitutes a promising alternative to GAN-based models for generating high-quality medical images, leading to improved diversity and less artifacts in the generated images.

Recent advances in computer vision have shown promising results in image generation. Diffusion probabilistic models have generated realistic images from textual input, as demonstrated by DALL-E 2, Imagen, and Stable Diffusion. However, their use in medicine, where imaging data typically comprises three-dimensional volumes, has not been systematically evaluated. Synthetic images may play a crucial role in privacy-preserving artificial intelligence and can also be used to augment small datasets. We show that diffusion probabilistic models can synthesize high-quality medical data for magnetic resonance imaging (MRI) and computed tomography (CT). For quantitative evaluation, two radiologists rated the quality of the synthesized images regarding \"realistic image appearance\", \"anatomical correctness\", and \"consistency between slices\". Furthermore, we demonstrate that synthetic images can be used in self-supervised pre-training and improve the performance of breast segmentation models when data is scarce (Dice scores, 0.91 [without synthetic data], 0.95 [with synthetic data]).

Artificial Intelligence (AI) can support diagnostic workflows in oncology by aiding diagnosis and providing biomarkers directly from routine pathology slides. However, AI applications are vulnerable to adversarial attacks. Hence, it is essential to quantify and mitigate this risk before widespread clinical use. Here, we show that convolutional neural networks (CNNs) are highly susceptible to white- and black-box adversarial attacks in clinically relevant weakly-supervised classification tasks. Adversarially robust training and dual batch normalization (DBN) are possible mitigation strategies but require precise knowledge of the type of attack used in the inference. We demonstrate that vision transformers (ViTs) perform equally well compared to CNNs at baseline, but are orders of magnitude more robust to white- and black-box attacks. At a mechanistic level, we show that this is associated with a more robust latent representation of clinically relevant categories in ViTs compared to CNNs. Our results are in line with previous theoretical studies and provide empirical evidence that ViTs are robust learners in computational pathology. This implies that large-scale rollout of AI models in computational pathology should rely on ViTs rather than CNN-based classifiers to provide inherent protection against perturbation of the input data, especially adversarial attacks. Artificial Intelligence can support diagnostic workflows in oncology, but they are vulnerable to adversarial attacks. Here, the authors show that convolutional neural networks are highly susceptible to white- and black-box adversarial attacks in clinically relevant classification tasks.

Unmasking the decision making process of machine learning models is essential for implementing diagnostic support systems in clinical practice. Here, we demonstrate that adversarially trained models can significantly enhance the usability of pathology detection as compared to their standard counterparts. We let six experienced radiologists rate the interpretability of saliency maps in datasets of X-rays, computed tomography, and magnetic resonance imaging scans. Significant improvements are found for our adversarial models, which are further improved by the application of dual-batch normalization. Contrary to previous research on adversarially trained models, we find that accuracy of such models is equal to standard models, when sufficiently large datasets and dual batch norm training are used. To ensure transferability, we additionally validate our results on an external test set of 22,433 X-rays. These findings elucidate that different paths for adversarial and real images are needed during training to achieve state of the art results with superior clinical interpretability. Unmasking the decision making process of machine learning models is essential for implementing diagnostic support systems in clinical practice. Here, the authors demonstrate that adversarially trained models can significantly enhance the usability of pathology detection as compared to their standard counterparts.

Blueberry ( Vaccinium spp.) is a popular healthy fruit worldwide. The health value of blueberry is mainly because the fruit is rich in anthocyanins, which have a strong antioxidant capacity. However, because blueberry is a non-model plant, little is known about the structural and regulatory genes involved in anthocyanin synthesis in blueberries. Previous studies have found that spraying 1,000 mg/L abscisic acid at the late green stage of “Jersey” highbush blueberry fruits can increase the content of anthocyanins. In this experiment, the previous results were verified in “Brightwell” rabbiteye blueberry fruits. Based on the previous results, the anthocyanin accumulation process in blueberry can be divided into six stages from the late green stage to the mature stage, and the transcriptome was used to systematically analyze the blueberry anthocyanin synthesis process. Combined with data from previous studies on important transcription factors regulating anthocyanin synthesis in plants, phylogenetic trees were constructed to explore the key transcription factors during blueberry fruit ripening. The results showed that ABA increased the anthocyanin content of blueberry fruits during veraison. All structural genes and transcription factors (MYB, bHLH, and WD40) involved in the anthocyanin pathway were identified, and their spatiotemporal expression patterns were analyzed. The expression of CHS , CHI , DFR , and LDOX/ANS in ABA-treated fruits was higher in the last two stages of maturity, which was consistent with the change in the anthocyanin contents in fruits. In general, six MYB transcription factors, one bHLH transcription factor and four WD40 transcription factors were found to change significantly under treatment during fruit ripening. Among them, VcMYBA plays a major role in the regulation of anthocyanin synthesis in ABA signaling. This result preliminarily explained the mechanism by which ABA increases the anthocyanin content and improves the efficiency of the industrial use of blueberry anthocyanins.

Optical diffraction neural networks (ODNNs) represent a promising advancement in computational optics, with significant potential for applications in image classification, image reconstruction, and biomedical imaging. By using the principles of light diffraction for neural network computations, ODNNs enable low-power, real-time data processing without the need for traditional electronic computing units. This review provides an overview of the foundational concepts behind ODNNs, starting with the principles of artificial neurons and progressing to the specific implementation of optical diffraction in neural network architectures. We examine recent advancements in key components of ODNNs, including optical signal processing, activation functions, and training algorithms. Additionally, we highlight the practical applications of ODNNs in areas such as signal analysis, optical imaging, image processing, and high-dimensional optical communications. This paper concludes with a discussion of the current challenges and future directions for ODNN research, emphasizing the potential for overcoming existing limitations and further expanding their capabilities.

Predicting acute toxicity across species is essential for early-stage drug safety evaluation. While recent efforts have primarily focused on improving predictive accuracy, they often fail to address two critical issues: the substantial divergence in toxicity mechanisms among different species, and the inherent noise present in experimental data. To bridge this gap, we introduce a Probabilistic Multitask Active Learning (PMAL) framework for multi-species acute toxicity prediction. Our framework integrates two key modules: a Probabilistic Multitask Learning (PML) component which jointly models the predictive distributions of multiple toxicity endpoints from a probabilistic viewpoint, and an Uncertainty-based Active Learning (UAL) component which strategically selects the most informative compounds for experimental annotation based on predictive uncertainty. Empirical evaluations demonstrate that PMAL surpasses state-of-the-art methods and is capable of providing well-calibrated uncertainty estimates for small molecules across diverse toxicity endpoints. Beyond advancing multi-species toxicity prediction, the core design principles of PMAL offer a generalizable paradigm for learning in noisy multi-task environments.

TRIM protein family is an evolutionarily conserved gene family implicated in a number of critical processes including inflammation, immunity, antiviral and cancer. In an effort to profile the expression patterns of TRIM superfamily in several non-small cell lung cancer (NSCLC) cell lines, we found that the expression of 10 TRIM genes including TRIM3, TRIM7, TRIM14, TRIM16, TRIM21, TRIM22, TRIM29, TRIM59, TRIM66 and TRIM70 was significantly upregulated in NSCLC cell lines compared with the normal human bronchial epithelial (HBE) cell line, whereas the expression of 7 other TRIM genes including TRIM4, TRIM9, TRIM36, TRIM46, TRIM54, TRIM67 and TRIM76 was significantly down-regulated in NSCLC cell lines compared with that in HBE cells. As TRIM59 has been reported to act as a proto-oncogene that affects both Ras and RB signal pathways in prostate cancer models, we here focused on the role of TRIM59 in the regulation of NSCLC cell proliferation and migration. We reported that TRIM59 protein was significantly increased in various NSCLC cell lines. SiRNA-induced knocking down of TRIM59 significantly inhibited the proliferation and migration of NSCLC cell lines by arresting cell cycle in G2 phase. Moreover, TRIM59 knocking down affected the expression of a number of cell cycle proteins including CDC25C and CDK1. Finally, we knocked down TRIM59 and found that p53 protein expression levels did not upregulate, so we proposed that TRIM59 may promote NSCLC cell growth through other pathways but not the p53 signaling pathway.