Numerous studies have accelerated the knowledge expansion on the role of gut microbiota in inflammatory bowel disease (IBD). However, the precise mechanisms behind host-microbe cross-talk remain largely undefined, due to the complexity of the human intestinal ecosystem and multiple external factors. In this review, we introduce the interactome concept to systematically summarize how intestinal dysbiosis is involved in IBD pathogenesis in terms of microbial composition, functionality, genomic structure, transcriptional activity, and downstream proteins and metabolites. Meanwhile, this review also aims to present an updated overview of the relevant mechanisms, high-throughput multi-omics methodologies, different types of multi-omics cohort resources, and computational methods used to understand host-microbiota interactions in the context of IBD. Finally, we discuss the challenges pertaining to the integration of multi-omics data in order to reveal host-microbiota cross-talk and offer insights into relevant future research directions.

Multi-omics clustering has emerged as a powerful approach for understanding complex biological systems and enabling cancer subtyping by integrating diverse omics data. Existing methods primarily focus on the integration of different types of omics data, often overlooking the value of clinical context. In this study, we propose a novel framework that incorporates clinical features extracted from large language model (LLM) to enhance multi-omics clustering. Leveraging clinical data extracted from pathology reports using a BERT-based model, our framework converts unstructured medical text into structured clinical features. These features are integrated with omics data through an autoencoder, enriching the information content of each omics layer to improve feature extraction. The extracted features are then projected into a latent subspace using singular value decomposition (SVD), followed by spectral clustering to obtain the final clustering result. We evaluate the proposed framework on six cancer datasets on three omics levels, comparing it with several state-of-the-art methods. The experimental results demonstrate that the proposed framework outperforms existing methods in multi-omics clustering for cancer subtyping. Moreover, the results highlight the efficacy of integrating clinical features derived from LLM, significantly enhancing clustering performance. This work underscores the importance of clinical context in multi-omics analysis and showcases the transformative potential of LLM in advancing precision medicine.

It is known that microbiome and health are related, in addition, recent research has found that microbiome has potential clinical uses. These facts highlight the importance of the microbiome in actual science. However, microbiome data has some characteristics that makes its statistical study challenging. In recent years, longitudinal and non-longitudinal methods have been designed to analyze the microbiota and knowing more about the bacterial behavior. In this article in the form of a review we summarize the characteristics of microbiome data and the statistical methods most widespread to analyze it. We have taken into account if the strategies are longitudinal or not. We also classify the methods based on their specific analytical objectives and based on their mathematical characteristics. The methods are structured according to their biological goals and mathematical features, ensuring that the insights provided are both relevant and accessible to professionals in biology and statistics. We present this review as a reference for the most widely used methods in microbiome data analysis and as a foundation for identifying potential areas for future research. We want to point out that this review can be particularly useful to remark the importance of the methodology designed in order to study microbiome longitudinal datasets.

The purpose of integrating different omics data is to study cellular heterogeneity at the level of transcriptional regulation from different gene levels, which can effectively identify cell types and reveal the pathogenesis of Alzheimer's disease (AD) from two perspectives. However, implementing such algorithms faces challenges such as high data noise levels, increased dimensionality, and computational complexity. In this study, multigraph regularization constraints were introduced in the network-based integrative clustering algorithm (MGR-NIC) to remove redundant features and keep the geometry structures underlying the data by fusing two types of data (snRNA-seq and snATAC-seq) of glial cells from AD samples. The effectiveness of the MGR-NIC algorithm was validated using both simulation datasets and real datasets derived from various tissues. The MGR-NIC algorithm can improve clustering accuracy by selecting features that better represent the dataset's structure. The clustering results obtained with the MGR-NIC algorithm show strong consistency with the clustering results inherent to the published DLPFC dataset, while the classification results generated using the NIC algorithm often lead to cluster overlap when applied to the DLPFC dataset. We will use the same state-of-the-art algorithms for a comprehensive evaluation with our proposed MGR-NIC algorithm, including NIC, scAI, Multi-Omics Factor Analysis v2, and JSNMF. MGR-NIC is the most stable and reliable method, implying its robustness across different datasets and its reliability in yielding consistent and accurate results.

Technological advances have enabled us to profile multiple omics layers with spatial information, significantly enhancing spatial domain detection and advancing a variety of biomedical research fields. Despite these advancements, there is a notable lack of effective methods for modeling spatial multi-omics data. We introduce SMOPCA, a Spatial Multi-Omics Principal Component Analysis method designed to perform joint dimension reduction on multimodal data while preserving spatial dependencies. Extensive experiments reveal that SMOPCA outperforms existing single-modal and multimodal dimension reduction and clustering methods, across both single-cell and spatial multi-omics datasets derived from diverse technologies and tissue structures.

Research into plant gene function is crucial for developing strategies to increase crop yields. The recent introduction of large language models (LLMs) offers a means to aggregate large amounts of data into a queryable format, but the output can contain inaccurate or false claims known as hallucinations. To minimize such hallucinations and produce high-quality knowledge-based outputs, the abstracts of over 60 000 plant research articles are compiled into a Chroma database for retrieval-augmented generation (RAG). Then linguistic data are used from 13 993 Arabidopsis (Arabidopsis thaliana) phenotypes and 23 323 gene functions to fine-tune the LLM Llama3-8B, producing PlantGPT, a virtual expert in Arabidopsis phenotype-gene research. By evaluating answers to test questions, it is demonstrated that PlantGPT outperforms general LLMs in answering specialized questions. The findings provide a blueprint for functional genomics research in food crops and demonstrate the potential for developing LLMs for plant research modalities. To provide broader access and facilitate adoption, the online tool http://www.plantgpt.icu is developed, which will allow researchers to use PlantGPT in their scientific investigations.

Multi-omics profiling integrates genomic, epigenomic, transcriptomic, and proteomic data, essential for understanding complex health and disease pathways. This review highlights the transformative potential of combining optical nanosensors with artificial intelligence (AI). It is possible to identify disease-specific biomarkers using real-time and sensitive molecular interactions. These technologies are precious for genetic, epigenetic, and proteomic changes critical to disease progression and treatment response. AI improves multi-omics profiling by analyzing large, diverse data sets and common patterns traditional methods overlook. Machine learning tools Biomarkers Discovery is revolutionizing, drug resistance is being understood, and medicine is being personalized as the combination of AI and nanosensors has advanced the detection of DNA methylation and proteomic signatures and improved our understanding of cancer, cardiovascular disease and vascular disease. Despite these advances, challenges still exist. Difficulties in integrating data sets, retaining sensors, and building scalable computing tools are the biggest obstacles. It also examines various solutions with advanced AI algorithms and innovations, including fabrication in nanosensor design. Moreover, it highlights the potential of nanosensor-assisted, AI-driven multi-omics profiling to revolutionize disease diagnosis and treatment. As technology advances, these tools pave the way for faster diagnosis, more accurate treatment and improved patient outcomes, offering new hope for personalized medicine.

The human solute carrier (SLC) superfamily of ~460 membrane transporters remains the largest understudied protein family despite its therapeutic potential. To advance SLC research, we developed a comprehensive knowledgebase that integrates systematic multi-omics data sets with selected curated information from public sources. We annotated SLC substrates through literature curation, compiled SLC disease associations using data mining techniques, and determined the subcellular localization of SLCs by combining annotations from public databases with an immunofluorescence imaging approach. This SLC-centric knowledge is made accessible to the scientific community via a web portal featuring interactive dashboards and visualization tools. Utilizing this systematically collected and curated resource, we computationally derived an integrated functional landscape for the entire human SLC superfamily. We identified clusters with distinct properties and established functional distances between transporters. Based on all available data sets and their integration, we assigned biochemical/biological functions to each SLC, making this study one of the largest systematic annotations of human gene function and a potential blueprint for future research endeavors.

Hepatocellular carcinoma (HCC) is the most prevalent type of liver cancer and a leading cause of cancer-related deaths globally. The tumour microenvironment (TME) influences treatment response and prognosis, yet its heterogeneity remains unclear.

The unsupervised machine learning methods- agglomerative hierarchical clustering, Multi-Omics Factor Analysis with K-means++, and an autoencoder with K-means++ - stratified patients using microarray data from HCC samples. Immune deconvolution algorithms estimated the proportions of infiltrating immune cells across identified clusters.

Thirteen genes were found to influence HCC subtyping in both primary and validation datasets, with three genes-TOP2A, DCN, and MT1E-showing significant associations with survival and recurrence. DCN, a known tumour suppressor, was significant across datasets and associated with improved survival, potentially by modulating the TME and promoting an anti-tumour immune response.

The discovery of the 13 conserved genes is an important step toward understanding HCC heterogeneity and the TME, potentially leading to the identification of more reliable biomarkers and therapeutic targets. We have stratified and validated the liver cancer populations. The findings suggest further research is needed to explore additional factors influencing the TME beyond gene expression, such as tumour microbiome and stromal cell interactions.

The rapid growth of multi-omics datasets and the wealth of biological knowledge necessitates the development of effective methods for their integration. Such methods are essential for building predictive models and identifying drug targets based on a limited number of samples. We propose a framework called GNNRAI for the supervised integration of multi-omics data with biological priors represented as knowledge graphs. Our framework leverages graph neural networks (GNNs) to model the correlation structures among features from high-dimensional 'omics data, which reduces the effective dimensions in data and enables us to analyze thousands of genes simultaneously using hundreds of samples. Furthermore, our framework incorporates explainability methods to elucidate informative biomarkers. We apply our framework to Alzheimer's disease (AD) multi-omics data, showing that the integration of transcriptomics and proteomics data with prior AD knowledge is effective, improving the prediction accuracy of AD status over single-omics analyses and highlighting both known and novel AD-predictive biomarkers.

Urbanization affects environmental exposures and lifestyle, shaping immune system variation and influencing disease susceptibility and vaccine responses. Here, we present systems analysis of immune profiles across the rural-urban gradient, comparing rural and urban Senegalese with urban Dutch individuals. By integrating single-cell phenotyping, metabolic profiling, and functional analysis, we reveal a trajectory of immune remodeling along the gradient. This includes enrichment of proinflammatory CD11c+ B cells associated with altered IgG Fc glycosylation, adaptive NK cells with reduced responsiveness to accessory cytokines, and CD161+CD4+T cells with enhanced cytokine production in rural settings. Metabolic perturbation studies demonstrated distinct dependencies on glycolysis, pentose phosphate pathway, and fatty acid synthesis for cellular cytokine responses across populations. We validate core rural-urban immune signatures in an independent Indonesian cohort, suggesting shared immunological adaptations to urbanization across ancestries and geographical areas. Our findings provide insights into rural-urban immune function in understudied populations.

Alzheimer's disease (AD) is a complex and progressive neurodegenerative disorder, characterized by diverse cognitive and functional impairments that manifest heterogeneously across individuals, domains, and time. The accurate assessment of AD's severity and progression requires integrating a variety of data modalities, including multivariate longitudinal neuropsychological tests and multi-omics datasets such as metabolomics and lipidomics. These data sources provide valuable insights into risk factors associated with dementia onset. However, effectively utilizing omics data in dynamic risk estimation for AD progression is challenging due to issues including high dimensionality, heterogeneity, and complex intercorrelations. To address these challenges, we develop a novel joint-modeling framework that effectively combines multi-omics factor analysis (MOFA) for dimension reduction and feature extraction with a multivariate functional mixed model (MFMM) for modeling longitudinal outcomes. This integrative joint modeling approach enables dynamic evaluation of dementia risk by leveraging both omics and longitudinal data. We validate the efficacy of our integrative model through extensive simulation studies and its practical application to the Alzheimer's Disease Neuroimaging Initiative (ADNI) dataset.

The integration and analysis of multi-modal data are increasingly essential across various domains including bioinformatics. As the volume and complexity of such data grow, there is a pressing need for computational models that not only integrate diverse modalities but also leverage their complementary information to improve clustering accuracy and insights, especially when dealing with partial observations with missing data.

We propose Generalized Probabilistic Canonical Correlation Analysis (GPCCA), an unsupervised method for the integration and joint dimensionality reduction of multi-modal data. GPCCA addresses key challenges in multi-modal data analysis by handling missing values within the model, enabling the integration of more than two modalities, and identifying informative features while accounting for correlations within individual modalities. The model demonstrates robustness to various missing data patterns and provides low-dimensional embeddings that facilitate downstream clustering and analysis. In a range of simulation settings, GPCCA outperforms existing methods in capturing essential patterns across modalities. Additionally, we demonstrate its applicability to multi-omics data from TCGA cancer datasets and a multi-view image dataset.

GPCCA offers a useful framework for multi-modal data integration, effectively handling missing data and providing informative low-dimensional embeddings. Its performance across cancer genomics and multi-view image data highlights its robustness and potential for broad application. To make the method accessible to the wider research community, we have released an R package, GPCCA, which is available at https://github.com/Kaversoniano/GPCCA.

Alzheimer's disease (AD) is a complex neurodegenerative disorder with an unclear pathogenesis; the traditional ″single gene-single target-single drug″ strategy is insufficient for effective treatment. This study explores a novel strategy for the multitarget therapy of AD by integrating multiomics data and employing network analysis. Different from conventional single-target methods, TCnet adopts a mechanism-driven strategy, utilizing multiomics data to decompose disease mechanisms, construct potential target combinations, and prioritize the optimal combinations using a scoring function. TCnet not only advances our understanding of disease mechanisms but also facilitates large-scale drug screening. This approach was further employed to screen active compounds from Huang-Lian-Jie-Du-Tang (HLJDT), identifying quercetin as a candidate targeting GSK3β and ADAM17. Subsequent in vitro experiments confirmed the neuroprotective and anti-inflammatory effects of quercetin. Overall, TCnet offers a promising approach for predicting target combinations and provides new insights and directions for drug discovery in AD.

Systems biology is a holistic approach to biological sciences that combines experimental and computational strategies, aimed at integrating information from different scales of biological processes to unravel pathophysiological mechanisms and behaviours. In this scenario, high-throughput technologies have been playing a major role in providing huge amounts of omics data, whose integration would offer unprecedented possibilities in gaining insights on diseases and identifying potential biomarkers. In the present review, we focus on strategies that have been applied in literature to integrate genomics, transcriptomics, proteomics, and metabolomics in the year range 2018-2024. Integration approaches were divided into three main categories: statistical-based approaches, multivariate methods, and machine learning/artificial intelligence techniques. Among them, statistical approaches (mainly based on correlation) were the ones with a slightly higher prevalence, followed by multivariate approaches, and machine learning techniques. Integrating multiple biological layers has shown great potential in uncovering molecular mechanisms, identifying putative biomarkers, and aid classification, most of the time resulting in better performances when compared to single omics analyses. However, significant challenges remain. The high-throughput nature of omics platforms introduces issues such as variable data quality, missing values, collinearity, and dimensionality. These challenges further increase when combining multiple omics datasets, as the complexity and heterogeneity of the data increase with integration. We report different strategies that have been found in literature to cope with these challenges, but some open issues still remain and should be addressed to disclose the full potential of omics integration.

Structural and functional alterations in the brain's reward circuitry are present in cocaine use disorder (CocUD), but their molecular underpinnings remain unclear. To investigate these mechanisms, we performed single-nuclei multiome profiling on postmortem caudate nucleus tissue from six individuals with CocUD and eight controls. We profiled 30,030 nuclei, identifying 13 cell types including D1- and D2-medium spiny neurons (MSNs) and glial cells. We observed 1485 differentially regulated genes and 10,342 differentially accessible peaks, with alterations in MSNs and astrocytes related to neurotransmitter activity and synapse organization. Gene regulatory network analysis identified transcription factors including ZEB1 as exhibiting distinct CocUD-specific subclusters, activating downstream expression of ion- and calcium-channels in MSNs. Further, PDE10A emerged as a potential drug target, showing conserved effects in a rat model. This study highlights cell type-specific molecular alterations in CocUD and provides targets for further investigation, demonstrating the value of multi-omics approaches in addiction research.

Malignant digestive tract tumors are highly prevalent and fatal tumor types globally, often diagnosed at advanced stages due to atypical early symptoms, causing patients to miss optimal treatment opportunities. Traditional endoscopic and pathological diagnostic processes are highly dependent on expert experience, facing problems such as high misdiagnosis rates and significant inter-observer variations. With the development of artificial intelligence (AI) technologies such as deep learning, real-time lesion detection with endoscopic assistance and automated pathological image analysis have shown potential in improving diagnostic accuracy and efficiency. However, relevant applications still face challenges including insufficient data standardization, inadequate interpretability, and weak clinical validation.

This study aims to systematically review the current applications of artificial intelligence in diagnosing malignant digestive tract tumors, focusing on the progress and bottlenecks in two key areas: endoscopic examination and pathological diagnosis, and to provide feasible ideas and suggestions for subsequent research and clinical translation.

A systematic literature search strategy was adopted to screen relevant studies published between 2017 and 2024 from databases including PubMed, Web of Science, Scopus, and IEEE Xplore, supplemented with searches of early classical literature. Inclusion criteria included studies on malignant digestive tract tumors such as esophageal cancer, gastric cancer, or colorectal cancer, involving the application of artificial intelligence technology in endoscopic diagnosis or pathological analysis. The effects and main limitations of AI diagnosis were summarized through comprehensive analysis of research design, algorithmic methods, and experimental results from relevant literature.

In the field of endoscopy, multiple deep learning models have significantly improved detection rates in real-time polyp detection, early gastric cancer, and esophageal cancer screening, with some commercialized systems successfully entering clinical trials. However, the scale and quality of data across different studies vary widely, and the generalizability of models to multi-center, multi-device environments remains to be verified. In pathological analysis, using convolutional neural networks, multimodal pre-training models, etc., automatic tissue segmentation, tumor grading, and assisted diagnosis can be achieved, showing good scalability in interactive question-answering. Nevertheless, clinical implementation still faces obstacles such as non-uniform data standards, lack of large-scale prospective validation, and insufficient model interpretability and continuous learning mechanisms.

Artificial intelligence provides new technological opportunities for endoscopic and pathological diagnosis of malignant digestive tract tumors, achieving positive results in early lesion identification and assisted decision-making. However, to achieve the transition from research to widespread clinical application, data standardization, model reliability, and interpretability still need to be improved through multi-center joint research, and a complete regulatory and ethical system needs to be established. In the future, artificial intelligence will play a more important role in the standardization and precision management of diagnosis and treatment of digestive tract tumors.

High-throughput multi-omic molecular profiling allows the probing of biological systems at unprecedented resolution. However, integrating and interpreting high-dimensional, sparse, and noisy multimodal datasets remains challenging. Deriving new biological insights with current methods is difficult because they are not rooted in biological principles but prioritise tasks like dimensionality reduction. Here, we introduce a framework that combines archetypal analysis, an approach grounded in biological principles, with deep learning. Using archetypes based on evolutionary trade-offs and Pareto optimality, MIDAA finds extreme data points that define the geometry of the latent space, preserving the complexity of biological interactions while retaining an interpretable output. We demonstrate that these extreme points represent cellular programmes reflecting the underlying biology. Moreover, we show that, compared to alternative methods, MIDAA can identify parsimonious, interpretable, and biologically relevant patterns from real and simulated multi-omics.

Chemotherapy-induced hepatotoxicity (CIH) is a significant concern in colorectal cancer (CRC) patients treated with the CAPEOX (capecitabine and oxaliplatin) regimen. Identifying predictive factors for CIH is crucial for clinical management.

This study analyzed colorectal tissue (CRT), plasma, and urine samples from CRC patients. Differentially expressed metabolites (DEMs) across these tissues were integrated for multi-omics analysis, and predictive models for CIH susceptibility were developed. An independent set of 75 plasma samples was used for validation.

A total of 492 differentially expressed compounds were identified in samples from 63 CRC patients, including 105, 149, and 238 DEMs in CRT, plasma, and urine, respectively. Lipids and lipid-like molecules were predominant in all samples. Among these, urine samples exhibited the highest variability and provided the strongest predictive power for CIH susceptibility. Principal component analysis (PCA) effectively differentiated normal patients from those with CIH. The study revealed steatosis as the primary pathological feature of CIH, with disrupted lipid metabolism emerging as a key characteristic. Predictive models constructed from multi-tissue metabolites profile exhibited high accuracy, with the plasma model achieving an AUC of 0.933 in external validation set. Our study underscores the importance of individual metabolic variations in CIH susceptibility, reflecting the complex interplay of genetic, environmental, and lifestyle factors.

This study emphasizes the critical role of alterations in lipid, polyamine, and purine metabolism, as well as impaired tissue repair mechanisms, were identified as key endogenous factors underlying CIH susceptibility. The developed predictive models demonstrate potential for clinical application in assessing CIH risk in CRC patients undergoing CAPEOX chemotherapy.

A holistic view on biological systems is today a reality with the application of multi-omic technologies. These technologies allow the profiling of genome, epigenome, transcriptome, proteome, metabolome as well as newly emerging 'omes.' While the multiple layers of data accumulate, their integration and reconciliation in a single system map is a cumbersome exercise that faces many challenges. Application to human health and disease requires large sample sizes, robust methodologies and high-quality standards.

We review the different methods used to integrate multi-omics, as recent ones including artificial intelligence. With proteomics as an anchor technology, we then present selected applications of its data combination with other omics layers in clinical research, mainly covering literature from the last five years in the Scopus and/or PubMed databases.

Multi-omics is powerful to comprehensively type molecular layers and link them to phenotype. Yet, technologies and data are very diverse and still strategies and methodologies to properly integrate these modalities are needed.

Chronic postoperative pain is present in approximately 20% of patients undergoing total knee arthroplasty. Studies indicate that pain mechanisms are associated with development and maintenance of chronic postoperative pain. The current study assessed pain sensitivity, inflammation, microRNAs, and psychological factors and combined these in a network to describe chronic postoperative pain. This study involved 75 patients with and without chronic postoperative pain after total knee arthroplasty. Clinical pain intensity, Oxford Knee Score, and pain catastrophizing were assessed as clinical parameters. Quantitative sensory testing was assessed to evaluate pain sensitivity and microRNAs, and inflammatory markers were likewise analyzed. Supervised multivariate data analysis with "Data Integration Analysis for Biomarker Discovery" using Latent cOmponents (DIABLO) was used to describe the chronic postoperative pain intensity. Two DIABLO models were constructed by dividing the patients into 3 groups or 2 defined by clinical pain intensities. Data Integration Analysis for Biomarker discovery using Latent cOmponents model explained chronic postoperative pain and identified factors involved in pain mechanistic networks among assessments included in the analysis. Developing models of 3 or 2 patient groups using the assessments and the networks could explain 81% and 69% of the variability in clinical postoperative pain intensity. The reduction of the number of parameters stabilized the models and reduced the explanatory value to 69% and 51%. This is the first study to use the DIABLO model for chronic postoperative pain and to demonstrate how different pain mechanisms form a pain mechanistic network. The complex model explained 81% of the variability of clinical pain intensity, whereas the less complex model explained 51% of the variability of clinical pain intensity.

The use of biomarkers in breast cancer has significantly improved patient outcomes through targeted therapies, such as hormone therapy anti-Her2 therapy and CDK4/6 or PARP inhibitors. However, existing knowledge does not fully encompass the diverse nature of breast cancer, particularly in triple-negative tumors. The integration of multi-omics and multimodal data has the potential to provide new insights into biological processes, to improve breast cancer patient stratification, enhance prognosis and response prediction, and identify new biomarkers. This review presents a comprehensive overview of the state-of-the-art multimodal (including molecular and image) data integration algorithms developed and with applicability to breast cancer stratification, prognosis, or biomarker identification. We examined the primary challenges and opportunities of these multimodal data integration algorithms, including their advantages, limitations, and critical considerations for future research. We aimed to describe models that are not only academically and preclinically relevant, but also applicable to clinical settings.

Hematopoietic stem cell transplantation (HCT) is potentially curative for numerous malignant and non-malignant diseases but can lead to lung injury due to chemoradiation toxicity, infection, and immune dysregulation. Bronchoalveolar lavage (BAL) is the most commonly used procedure for diagnostic sampling of the lung but is invasive, cannot be performed in medically fragile patients, and is challenging to perform serially. We previously showed that BAL transcriptomes representing pulmonary inflammation and cellular injury can phenotype post-HCT lung injury and predict mortality outcomes. However, whether peripheral blood testing is a suitable minimally-invasive surrogate for pulmonary sampling in the HCT population remains unknown. To address this question, we compared 210 paired BAL and peripheral blood transcriptomes obtained from 166 pediatric HCT patients at 27 children's hospitals. BAL and blood mRNA abundance showed minimal overall correlation at the level of individual genes, gene set enrichment scores, imputed cell fractions, and T- and B-cell receptor clonotypes. Instead, we identified significant site-specific transcriptional programs. In BAL, expression of innate and adaptive immune pathways was tightly co-regulated with expression of epithelial mesenchymal transition and hypoxia pathways, and these signatures were associated with mortality. In contrast, in blood, expression of endothelial injury, DNA repair, and cellular metabolism pathways was associated with mortality. Integration of paired BAL and blood transcriptomes dichotomized patients into two groups, of which one group showed twice the rate of hypoxia and significantly worse outcomes within 7 days of enrollment. These findings reveal a compartmentalized injury response, where BAL and peripheral blood transcriptomes provide distinct but complementary insights into local and systemic mechanisms of post-HCT lung injury.

Drug repurposing identifies new therapeutic uses for the existing drugs originally developed for different indications, aiming at capitalizing on the established safety and efficacy profiles of known drugs. Thus, it is beneficial to bypass of early stages of drug development, and to reduction of the time and cost associated with bringing new therapies to market. Traditional experimental methods are often time-consuming and expensive, making artificial intelligence (AI) a promising alternative due to its lower cost, computational advantages, and ability to uncover hidden patterns. This review focuses on the availability of AI algorithms in drug development, and their positive and specific roles in revealing repurposing of the existing drugs, especially being integrated with virtual screening. It is shown that the existing AI algorithms excel at analyzing large-scale datasets, identifying the complicated patterns of drug responses from these datasets, and making predictions for potential drug repurposing. Building on these insights, challenges remain in developing efficient AI algorithms and future research, including integrating drug-related data across databases for better repurposing, enhancing AI computational efficiency, and advancing personalized medicine.

Rust fungi are highly destructive pathogens that pose a significant threat to crop production worldwide, especially cereals. Obligate biotrophy and, in many cases, complex life cycles make rust fungi particularly challenging to study. However, recent rapid advances in sequencing technologies and genomic analysis tools have revolutionised rust fungal research. It is anticipated that the increasing availability and ongoing substantial improvements in genome assemblies will propel the field of rust biology into the post-genomic era, instigating a cascade of research endeavours encompassing multi-omics and gene discoveries. This is especially the case for many cereal rust pathogens, for which continental-scale studies of virulence have been conducted over many years and historical collections of viable isolates have been sequenced and assembled. Genomic analysis plays a crucial role in uncovering the underlying causes of the high variability of virulence and the complexity of population dynamics in rust fungi. Here, we provide an overview of progress in rust genomics, discuss the strategies employed in genomic analysis, and elucidate the strides that will drive cereal rust biology into the post-genomic era.

Synechococcus elongatus PCC 7942 is a model organism for studying circadian regulation and bioproduction, where precise temporal control of metabolism significantly impacts photosynthetic efficiency and CO2-to-bioproduct conversion. Despite extensive research on core clock components, our understanding of the broader regulatory network orchestrating genome-wide metabolic transitions remains incomplete. We address this gap by applying machine learning tools and network analysis to investigate the transcriptional architecture governing circadian-controlled gene expression. While our approach showed moderate accuracy in predicting individual transcription factor-gene interactions - a common challenge with real expression data - network-level topological analysis successfully revealed the organizational principles of circadian regulation. Our analysis identified distinct regulatory modules coordinating day-night metabolic transitions, with photosynthesis and carbon/nitrogen metabolism controlled by day-phase regulators, while nighttime modules orchestrate glycogen mobilization and redox metabolism. Through network centrality analysis, we identified potentially significant but previously understudied transcriptional regulators: HimA as a putative DNA architecture regulator, and TetR and SrrB as potential coordinators of nighttime metabolism, working alongside established global regulators RpaA and RpaB. This work demonstrates how network-level analysis can extract biologically meaningful insights despite limitations in predicting direct regulatory interactions. The regulatory principles uncovered here advance our understanding of how cyanobacteria coordinate complex metabolic transitions and may inform metabolic engineering strategies for enhanced photosynthetic bioproduction from CO2.

Type 1 diabetes (T1D) is linked to an altered gut microbiota characterized by reduced short-chain fatty acid (SCFA) production. Oral delivery of a SCFA-yielding biotherapy in adults with T1D was followed by increased SCFAs, altered gut microbiota and immunoregulation, as well as delaying diabetes in preclinical models. Here, we show that SCFA-biotherapy in humans is accompanied by remodeling of the gut proteome and mucosal immune homeostasis. Metabolomics showed arginine, glutamate, nucleotide and tryptophan metabolism were enriched following the SCFA-biotherapy, and found metabolites that correlated with glycemic control. Fecal microbiota transfer demonstrated that the microbiota of SCFA-responders delayed diabetes progression in humanized gnotobiotic mice. The protected mice increased similar metabolite pathways to the humans including producing aryl-hydrocarbon receptor ligands and reducing inflammatory mucosal immunity and increasing IgA production in the gut. These data demonstrate that a potent SCFA immunomodulator promotes multiple beneficial pathways and supports targeting the microbiota as an approach against T1D. Trial registration: Australia New Zealand Clinical Trials Registry ACTRN12618001391268.

The tumor microenvironment (TME) of triple-negative breast cancer (TNBC) is a highly heterogeneous and very aggressive form of the disease that has few suitable treatment options; however, spatial transcriptomics (ST) is a powerful tool for elucidation of the TME in TNBC. Because of its spatial context preservation, ST has a unique capability to map tumor-stroma interactions, immune infiltration, and therapy resistance mechanisms (which are key to understanding TNBC progression), compared with conventional transcriptomics. This review shows the use of ST in TNBC, its utilization in spatial biomarker identification, intratumoral heterogeneity definition, molecular subtyping refinement, and prediction of immunotherapy responses. Recent insight from ST-driven insights has explained the key spatial patterns on immune evasion, chemotherapy resistance, racial disparities of TNBC, and aspects for patient stratification and therapeutic decision. With the integration of ST with the subjects of proteomics and imaging mass cytometry, this approach has been enlarged and is now applied in precision medicine and biomarker discovery. Recently, advancements in AI-based spatial analysis for tumor classification, identification of biomarkers, and creation of therapy prediction models have occurred. However, continued developments in ST technologies, computational tools, and partnerships amongst multiple centers to facilitate the integration of ST into clinical routine practice are needed to unlock novel therapeutic targets.

The continuous advancement of single-cell multimodal omics (scMulti-omics) technologies offers unprecedented opportunities to measure various modalities, including RNA expression, protein abundance, gene perturbation, DNA methylation, and chromatin accessibility at single-cell resolution. These advances hold significant potential for breakthroughs by integrating diverse omics modalities. However, the data generated from different omics layers often face challenges due to high dimensionality, heterogeneity, and sparsity, which can adversely impact the accuracy and efficiency of data integration analyses. To address these challenges, we propose a high-precision analysis method called scMGAT (single-cell multiomics data analysis based on multihead graph attention networks). This method effectively coordinates reliable information across multiomics data sets using a multihead attention mechanism, allowing for better management of the heterogeneous characteristics inherent in scMulti-omics data. We evaluated scMGAT's performance on eight sets of real scMulti-omics data, including samples from both human and mouse. The experimental results demonstrate that scMGAT significantly enhances the quality of multiomics data and improves the accuracy of cell-type annotation compared to state-of-the-art methods. scMGAT is now freely accessible at https://github.com/Xingyu-Liao/scMGAT.

Despite significant efforts to understand pathophysiological processes underlying cardiac diseases, the molecular causes for the most part remain unresolved. Rapid advancements in -omics technologies, and their application in cardiac research, offer new insight into cardiac remodeling in disease states. This review aims to provide an accessible overview of recent advances in omics approaches for studying cardiac remodeling, catering to readers without extensive prior expertise.

We provide a methodologically focused overview of current methods for performing transcriptomics and proteomics, including their extensions for single-cell and spatial measurements. We discuss approaches to integrate data across modalities, resolutions and time. Key recent applications within the cardiac field are highlighted. Each -omics modality can provide insight, yet each existing experimental method has technical or conceptual limitations. Integrating data across multiple modalities can leverage strengths and mitigate weaknesses, ultimately enhancing our understanding of cardiac pathophysiology.

Recent development of single-cell technology across multiple omics platforms has provided new ways to obtain holistic views of cells to study disease pathobiology. Alzheimer's disease (AD) is the most common form of dementia worldwide, yet the detailed understanding of its cellular and molecular mechanisms remains limited. In this study, we analyzed paired single-cell transcriptomic (scRNA-seq) and chromatin accessibility (scATAC-seq) data from the Seattle Alzheimer's Disease Brain Cell Atlas (SEA-AD) Consortium to investigate the molecular mechanisms of AD at a cell-subpopulation-specific resolution focusing on glial cells. We benchmarked various multi-omics integration methods using diverse metrics and built an analytic workflow that enabled effective batch correction and cross-modality alignment, creating a unified cell state space. Through integrative analysis of 26 human brain samples, we uncovered AD-associated gene expression and pathway changes in glial subpopulations and highlighted important transcriptomic and epigenomic signatures via functional inference and interpretable machine learning paradigms, discovering the profound involvement of the Solute Carrier proteins (SLC) family genes in multiple glial cell types. We also identified glial cell-specific regulatory programs mediated by key transcription factors such as JUN and FOSL2 in astrocytes, the Zinc Finger (ZNF) family genes in microglia, and the SOX family of transcription factors in oligodendrocytes. Our study provides a comprehensive workflow and a high-resolution view of how glial regulatory programs are disrupted in AD. Our findings offer novel insights into disease-related changes in gene regulation and suggest potential targets for further research and therapy.

Chronic kidney diseases (CKDs) are a global health concern, necessitating a comprehensive understanding of their complex pathophysiology. This study explores the use of 2 complementary multidimensional -omics data integration methods to elucidate mechanisms of CKD progression as a proof of concept. Baseline biosamples from 37 participants with CKD in the Clinical Phenotyping and Resource Biobank Core (C-PROBE) cohort with prospective longitudinal outcome data ascertained over 5 years were used to generate molecular profiles. Tissue transcriptomic, urine and plasma proteomic, and targeted urine metabolomic profiling were integrated using 2 orthogonal multi-omics data integration approaches, one unsupervised and the other supervised. Both integration methods identified 8 urinary proteins significantly associated with long-term outcomes, which were replicated in an adjusted survival model using 94 samples from an independent validation group in the same cohort. The 2 methods also identified 3 shared enriched pathways: the complement and coagulation cascades, cytokine-cytokine receptor interaction pathway, and the JAK/STAT signaling pathway. Use of different multiscalar data integration strategies on the same data enabled identification and prioritization of disease mechanisms associated with CKD progression. Approaches like this will be invaluable with the expansion of high-dimension data in kidney diseases.

Advances in single-cell multi-omics technologies have deepened our understanding of cancer biology by integrating genomic, transcriptomic, epigenomic, and proteomic data at single-cell resolution. These single-cell multi-omics technologies provide unprecedented insights into tumour heterogeneity, tumour microenvironment, and mechanisms of therapeutic resistance, enabling the development of precision medicine strategies. The emerging field of single-cell multi-omics in genomic medicine has improved patient outcomes. However, most clinical applications still depend on bulk genomic approaches, which fail to directly capture the genomic variations driving cellular heterogeneity. In this review, we explore the common single-cell multi-omics platforms and discuss key analytical steps for data integration. Furthermore, we highlight emerging knowledge in therapeutic resistance and immune evasion, and the potential of new therapeutic innovations informed by single-cell multi-omics. Finally, we discuss the future directions of the application of single-cell multi-omics technologies. By bridging the gap between technological advancements and clinical implementation, this review provides a roadmap for leveraging single-cell multi-omics to improve cancer treatment and patient outcomes.

Small intestinal neuroendocrine tumors (siNETs) are rare bowel tumors arising from malignant enteroendocrine cells, which normally regulate digestion throughout the intestine. Though infrequent, their incidence is rising through better diagnosis, fostering research into their origin and treatment. To date, siNETs are considered to be a single entity and are clinically treated as such. Here, by performing a multi-omics analysis of siNETs, we unveil four distinct molecular groups with strong clinical relevance and provide a resource to study their origin and clinical features. Transcriptomic, genetic and DNA methylation profiles identify two groups linked to distinct enteroendocrine differentiation patterns, another with a strong immune phenotype, and the last with mesenchymal properties. This latter subtype displays the worst prognosis and resistance to treatments in line with infiltration of cancer-associated fibroblasts. These data provide insights into the origin and diversity of these rare diseases, in the hope of improving clinical research into their management.

Precision medicine is still not considered as a standard of care in obesity treatment, despite a large heterogeneity in the metabolic phenotype of individuals with obesity. One of the strongest factors influencing the variability in metabolic disease risk is adipose tissue (AT) dysfunction; however, there is little understanding of the link between distinct cell populations, cell-type-specific transcriptional programs, and disease severity. Here, we generated a comprehensive cellular map of subcutaneous and visceral AT of individuals with metabolically healthy and unhealthy obesity. By combining single-nucleus RNA-sequencing data with bulk transcriptomics and clinical parameters, we identified that mesothelial cells, adipocytes, and adipocyte-progenitor cells exhibit the strongest correlation with metabolic disease. Furthermore, we uncovered cell-specific transcriptional programs, such as the transitioning of mesothelial cells to a mesenchymal phenotype, that are involved in uncoupling obesity from metabolic disease. Together, these findings provide valuable insights by revealing biological drivers of clinical endpoints.