Biomedical image segmentation plays a critical role in clinical applications such as disease diagnosis and surgical planning. While deep learning, especially U-Net and its variants, have achieved impressive performance in this domain, their success heavily relies on manual architectural design, which limits scalability and generalizability. Neural architecture search (NAS) offers a promising solution by automating model design, but existing NAS-based methods still suffer from high computational costs and limited architectural diversity. In this work, we propose a novel NAS framework, termed ranking-aware predictorassisted architecture design (RPA2D), for efficient and accurate biomedical image segmentation. RPA2D introduces a lightweight and expressive search space comprising diverse convolution and pooling operations tailored for medical images. Furthermore, we design a contrastive learning-based ranking-aware performance predictor that enables accurate architecture ranking under limited supervision. Extensive experiments on public biomedical datasets demonstrate that RPA2D consistently outperforms both handcrafted and NAS-based baselines in segmentation accuracy while significantly reducing search overhead.

Multimodal medical image classification (MMIC) refers to the process of extracting and combining information from various modalities to classify medical images, ultimately improving diagnostic accuracy. Most existing methods extract discriminative features from different modalities for specific tasks but often lack adaptability to other tasks. Additionally, they suffer from poor interpretability, which is critical in medical image analysis, as understanding the decision-making process is essential. Genetic programming (GP), particularly multitree GP, provides a flexible framework for evolving multimodal features. However, current multitree GP methods only perform feature-level fusion and remain underexplored in medical multimodal feature learning. To address these issues, this paper proposes a novel multitree GP method, Multimodal Feature GP (MFGP), to automatically extract informative feature vectors from different modalities. To fully utilize both modality-specific features and fused multimodal features, we integrate feature-level fusion and decision-level fusion strategies into our framework. The performance of the proposed method is evaluated on two distinct MMIC tasks, namely polyp classification and glaucoma classification, representing different medical scenarios. The results are compared with both single-modality and multimodal methods. Experimental results demonstrate that the proposed method significantly outperforms all single-modality approaches and most multimodal benchmark methods. Further analysis reveals that the evolved models can effectively capture the unique characteristics of different modalities.

SUMMARY: The rapid progresses of high-throughput sequencing technology-based omics and mass spectrometry-based proteomics, such as data-independent acquisition and its penetration to clinical studies have generated increasing number of proteomic datasets containing hundreds to thousands of samples. To analyze these quantitative proteomic datasets and other omics (e.g. transcriptomics and metabolomics) datasets more efficiently and conveniently, we present a web server-based software tool ProteomeExpert implemented in Docker, which offers various analysis tools for experimental design, data mining, interpretation and visualization of quantitative proteomic datasets. ProteomeExpert can be deployed on an operating system with Docker installed or with R language environment. AVAILABILITY AND IMPLEMENTATION: The Docker image of ProteomeExpert is freely available from https://hub.docker.com/r/lifeinfo/proteomeexpert. The source code of ProteomeExpert is also openly accessible at http://www.github.com/guomics-lab/ProteomeExpert/. In addition, a demo server is provided at https://proteomic.shinyapps.io/peserver/. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

Deep generative model, especially variational auto-encoder (VAE), has been successfully employed by more and more recommendation systems. The reason is that it combines the flexibility of probabilistic generative model with the powerful non-linear feature representation ability of deep neural networks. The existing VAE-based recommendation models are usually proposed under global assumption by incorporating simple priors, e.g., a single Gaussian, to regularize the latent variables. This strategy, however, is ineffective when the user is simultaneously interested in different kinds of items, i.e., the user’s preference may be highly diverse. In this paper, thus, we propose a Deep Global and Local Generative Model for recommendation to consider both local and global structure among users (DGLGM) under the Wasserstein auto-encoder framework. Besides keeping the global structure like the existing model, DGLGM adopts a non-parametric Mixture Gaussian distribution with several components to capture the diversity of the users’ preferences. Each component is corresponding to one local structure and its optimal size can be determined via the automatic relevance determination technique. These two parts can be seamlessly integrated and enhance each other. The proposed DGLGM can be efficiently inferred by minimizing its penalized upper bound with the aid of local variational optimization technique. Meanwhile, we theoretically analyze its generalization error bounds to guarantee its performance in sparse feedback data with diversity. By comparing with the state-of-the-art methods, the experimental results demonstrate that DGLGM consistently benefits the recommendation system in top-N recommendation task.

According to the sparseness of rating information, the quality of recommender systems has been greatly restricted. In order to solve this problem, much auxiliary information has been used, such as social networks, review information, and item description. Convolutional neural networks (CNNs) have been widely employed by recommender systems, it greatly improved the rating prediction's accuracy especially when combined with traditional recommendation methods. However, a large amount of research focuses on the consistency between the rating-based latent factor and review-based latent factor. But in fact, these two parts are completely different. In this article, the authors propose a model named collaboration matrix factorization (CMF) that combines a projection method with a convolutional matrix factorization (ConvMF) to extract the collaboration between rating-based latent factors and review-based latent factors that comes from the results of the CNN process. Extensive experiments on three real-world datasets show that the projection method achieves significant improvements over the existing baseline.