Shunyao Wu

Genomic Language Models (GLMs), leveraging their vast parameter scales and the similarities between DNA sequences and natural languages, demonstrate immense potential in processing large-scale genomic data and elucidating gene regulation and evolutionary relationships. However, the cross-species generalization capability of large genomic models has not yet been systematically evaluated. This study addresses this critical gap by benchmarking five GLMs (DNABERT-2, GROVER, HyenaDNA, NT-V2, and AgroNT) and a CNN baseline model using human (Homo sapiens) and rice (Oryza sativa) genomes across four downstream tasks: promoter detection, transcription start site (TSS) scanning, species classification, and gene region identification, through both zero-shot testing and fine-tuning. During testing, factors such as hyperparameters, early stopping protocols, and computational resources were fixed to ensure fairness, enabling us to systematically evaluate their performance and cross-species generalization capabilities. The results were further analyzed from multiple mathematical and representational perspectives to provide a more rigorous and objective assessment of each model’s performance. The results show that AgroNT consistently leads on rice tasks, while NT-V2 and DNABERT-2 achieved the best overall performance in fine-tuning and zero-shot experiments, respectively. Although their pretraining data did not include plants, they demonstrate excellent performance on rice-related tasks thanks to cross-species pretraining that enhances their generalization ability across human–rice domains. This benchmark study offers guidance on selecting appropriate genomic language models based on task characteristics and provides insights for future development in this field.

The human microbiome is crucial for health regulation and disease progression, presenting a valuable opportunity for health state classification. Traditional microbiome-based classification relies on pre-trained machine learning (ML) or deep learning (DL) models, which typically focus on microbial distribution patterns, neglecting the underlying relationships between microbes. As a result, model performance can be significantly affected by data sparsity, misclassified features, or incomplete microbial profiles. To overcome these challenges, we introduce Phylo-Spec, a phylogeny-driven deep learning algorithm that integrates multi-aspect microbial information for improved status recognition. Phylo-Spec fuses convolutional features of microbes within a phylogenetic hierarchy via a bottom-up iteration and significantly alleviates the challenges due to sparse data and inaccurate profiling. Additionally, the model dynamically assigns unclassified species to virtual nodes on the phylogenetic tree based on higher-level taxonomy, minimizing interferences from unclassified species. Phylo-Spec also captures the feature importance via an information gain-based mechanism through the phylogenetic structure propagation, enhancing the interpretability of classification decisions. Phylo-Spec demonstrated superior efficacy in microbiome status classification across two in silico synthetic data sets that simulate the aforementioned cases, outperforming existing ML and DL methods. Validation with real-world metagenomic and amplicon data further confirmed the model’s performance in multiple status classification, establishing a powerful framework for microbiome-based health state identification and microbe-disease association. The source code is available at https://github.com/qdu-bioinfo/Phylo-Spec . IMPORTANCE: The human microbiome profoundly influences health and disease, but current computational tools often overlook the evolutionary relationships among microbes, leading to incomplete or inaccurate interpretations of complex microbial data. Phylo-Spec provides a new way to understand the microbiome by combining microbial abundance, taxonomy, and phylogeny within a unified deep learning framework. This model not only improves the accuracy of health status classification but also highlights key microbial contributors linked to disease. By capturing both microbial diversity and evolutionary context, Phylo-Spec bridges the gap between bioinformatics and biological insight, offering a powerful and interpretable approach for advancing microbiome-based diagnostics and precision medicine.

The rapid development of large language models (LLMs) has revolutionized microbiome research by providing powerful tools for analyzing complex microbial data. Traditional methods often struggle with the volume and complexity of microbiome datasets, necessitating advanced computational techniques. LLMs, with their ability to model sequences and understand context, have enhanced microbial classification, functional prediction, and sequence analysis. These models also facilitate microbiome association mining, extracting meaningful microbe–disease and microbe–diet relationships from vast biomedical literature. In clinical diagnostics, LLMs support state recognition and disease detection by automating the extraction and integration of microbiome data. Despite their promise, challenges remain, including improving model interpretability, cross‐dataset generalization, and integrating multiomics data. This review highlights the transformative potential of LLMs in microbiome research, emphasizing their applications in sequence profiling, association mining, and clinical diagnostics, and discusses future directions for enhancing their effectiveness in precision medicine and ecological studies.

Background: Microbe-disease associations are integral to understanding complex diseases and their screening procedures. Objective: While numerous computational methods have been developed to detect these associations, their performance remains limited due to inadequate utilization of weighted inherent similarities and microbial taxonomy hierarchy. To address this limitation, we have introduced WTHMDA (weighted taxonomic heterogeneous network-based microbe-disease association), a novel deep learning framework. Methods: WTHMDA combines a weighted graph convolution network and the microbial taxonomy common tree to predict microbe-disease associations effectively. The framework extracts multiple microbe similarities from the taxonomy common tree, facilitating the construction of a microbe- disease heterogeneous interaction network. Utilizing a weighted DeepWalk algorithm, node embeddings in the network incorporate weight information from the similarities. Subsequently, a deep neural network (DNN) model accurately predicts microbe-disease associations based on this interaction network. Results: Extensive experiments on multiple datasets and case studies demonstrate WTHMDA's superiority over existing approaches, particularly in predicting unknown associations. Conclusion: Our proposed method offers a new strategy for discovering microbe-disease linkages, showcasing remarkable performance and enhancing the feasibility of identifying disease risk.

Background: The exploration of gene-disease associations is crucial for understanding the mechanisms underlying disease onset and progression, with significant implications for prevention and treatment strategies. Advances in high-throughput biotechnology have generated a wealth of data linking diseases to specific genes. While graph representation learning has recently introduced groundbreaking approaches for predicting novel associations, existing studies always overlooked the cumulative impact of functional modules such as protein complexes and the incompletion of some important data such as protein interactions, which limits the detection performance. Results: Addressing these limitations, here we introduce a deep learning framework called ModulePred for predicting disease-gene associations. ModulePred performs graph augmentation on the protein interaction network using L3 link prediction algorithms. It builds a heterogeneous module network by integrating disease-gene associations, protein complexes and augmented protein interactions, and develops a novel graph embedding for the heterogeneous module network. Subsequently, a graph neural network is constructed to learn node representations by collectively aggregating information from topological structure, and gene prioritization is carried out by the disease and gene embeddings obtained from the graph neural network. Experimental results underscore the superiority of ModulePred, showcasing the effectiveness of incorporating functional modules and graph augmentation in predicting disease-gene associations. This research introduces innovative ideas and directions, enhancing the understanding and prediction of gene-disease relationships.

The table recognition model rows and columns aggregated network (RCANet) uses a semantic segmentation approach to recognize table structure, and achieves better performance in table row and column segmentation. However, this model uses ResNet18 as the backbone network, and the model has 11.35 million parameters and a volume of 45.5 M, which is inconvenient to deploy to lightweight servers or mobile terminals. Therefore, from the perspective of model compression, this paper proposes the lightweight rows and columns attention aggregated network (LRCAANet), which uses the lightweight network ShuffleNetv2 to replace the original RCANet backbone network ResNet18 to simplify the model size. Considering that the lightweight network reduces the number of feature channels, it has a certain impact on the performance of the model. In order to strengthen the learning between feature channels, the rows attention aggregated (RAA) module and the columns attention aggregated (CAA) module are proposed. The RAA module and the CAA module add the squeeze and excitation (SE) module to the original row and column aggregated modules, respectively. Adding the SE module means the model can learn the correlation between channels and improve the prediction effect of the lightweight model. The experimental results show that our method greatly reduces the model parameters and model volume while ensuring low-performance loss. In the end, the average F1 score of our model is only 1.77% lower than the original model, the parameters are only 0.17 million, and the volume is only 0.8 M. Compared with the original model, the parameter amount and volume are reduced by more than 95%.

Everything is connected in the world. From small groups to global societies, the interactions among people, technology, and policies need sophisticated techniques to be perceived and forecasted. In social network, it has been concluded that the microblog users influence and microblog grade are nonlinearly dependent. However, to the best of our knowledge, the nonlinear influence predication of social network has not been explored in the existing literature. This article proposes a partial autoregression single index model to combine network structure (linear) and static covariates (nonparametric) flexibly. Compared with previous work, our model has fewer limits and more applications. The profile least squares estimation is employed to infer this semiparametric model, and variables selection is performed via the smoothly clipped absolute deviation penalty (SCAD). Simulations are conducted to demonstrate finite sample behaviors.

To address cold start problem by utilizing only rating information, this paper proposes an incremental group-specific framework for recommender systems. Firstly, a decoupled normalization method is introduced to extract preference patterns from ratings. Then, two incremental community detection methods are proposed to detect evolving communities in dynamic networks for capturing interest shifts, and new users/items corresponding communities according to the missing mechanism of ratings. Finally, an incremental group-specific model is proposed to incorporate evolving communities for recommender systems. A series of empirical analysis on three datasets is conducted to validate the rationality of grouping new users/items with missingness-related information. Experimental results show that the proposed framework can achieve better performance compared with other competitive methods, and is capable of handling the cold start problem and highly scalable with incremental data.