Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “omics”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 55 records · Page 3Linked to original sources

From molecular responses to environmental monitoring: advances and translational gaps in omics approaches in fish environmental toxicology.

Fish occupy a central position in aquatic ecosystems and serve as important bioindicators for environmental monitoring, as well as powerful translational models for understanding toxic mechanisms conserved across higher vertebrates. In recent years, omics techniques have proven to be powerful tools to address complex environmental questions that conventional toxicology methods cannot answer. Despite this potential, a critical translational gap remains between molecular findings and their use in ecological risk assessment frameworks. This review critically synthesizes advances across omics techniques including epigenomics, transcriptomics, metabolomics and proteomics and their integration. Special emphasis is placed on methodological considerations and practical aspects of these techniques in fish environmental toxicology and environmental monitoring. Evidence from single-omics studies suggests conserved biomarker signatures across species while characterizing complex phenomena like non-monotonic dose-response relationships, mixture toxicity and transgenerational and stereoselective effects with implications for population level monitoring. Multi-omics studies, especially those involving triple omics, further enhance mechanistic resolution by reconstructing adverse outcome pathways. We further evaluate using case studies when additional molecular layers provide critical insight and when they offer limited advantage, a strategic distinction with direct implications in environmental monitoring programmes. Finally, current limitations and future directions that will ultimately bridge the translational gap and hold promise for advancing mechanistic ecotoxicology and predictive environmental monitoring are discussed.

Animals↗

Omics in optic neuropathies: From molecular landscapes to personalized therapeutics.

Optic neuropathies comprise a heterogeneous group of disorders involving transient or permanent injury to retinal ganglion cells (RGCs) and their axons. Clinically, these neurodegenerative conditions manifest as dyschromatopsia, decreased visual acuity, and visual field defects, and in severe cases may ultimately lead to blindness and disability. The marked heterogeneity across disease subtypes, incompletely understood etiologies, and complex pathogenic mechanisms pose substantial challenges to precise diagnosis and effective treatment. Recent advances in omics technologies - including genomics, transcriptomics, proteomics, metabolomics, lipidomics, single-cell and spatial sequencing, and integrative multi-omics approaches - have ushered optic nerve degenerative disease research into an era of high-resolution comprehensive investigation. In this review, we summarize representative applications of omics approaches to elucidate genetic alterations, signaling dysregulation, metabolic reprogramming, and immune responses in optic neuropathies. We further discuss the emerging potential of multi-omics in identifying early diagnostic biomarkers and informing individualized therapeutic strategies. Finally, we provide a forward-looking perspective on the future trajectory of omics technologies and their prospects in both fundamental research and clinical translation, with the overarching aim of accelerating the bench-to-bedside transition in this critical eye disease field.

biomarkers↗

Improving recombinant protein productivity in CHO cells via multi-omics data integration.

Chinese hamster ovary (CHO) cells represent the dominant host system for the production of recombinant therapeutic proteins. In recent decades, extensive research has focused on process/media optimization and cell line engineering to improve both the productivity and quality of biopharmaceutical proteins produced in CHO cells. Nevertheless, the inherent complexity of biological pathways and the heterogeneous cellular responses to different environmental conditions have posed substantial challenges to traditional methodologies. Recent advances in omics technologies have enabled comprehensive characterization of CHO cell physiology, providing multidimensional molecular and phenotypic insights that facilitate the enhancement of recombinant protein production. This review first summarizes the methodologies and advances in CHO omics research, including genomics, transcriptomics, proteomics, metabolomics, and epigenomics. It then examines contemporary approaches to integrate and analyze multi-omics data in CHO cells. The review further elucidates how these multi-omics datasets can be strategically applied across various developmental stages, including cell line selection, genetic engineering, expression vector design, and bioprocess optimization. Finally, we explore the transformative potential of integrating multi-omics with artificial intelligence and discuss promising future research directions in CHO cell studies. These emerging paradigms offer novel opportunities for data-driven cell engineering and bioprocess optimization in CHO-based biomanufacturing.

Bioprocessing↗

Research progress and application prospects of multi-omics integration strategies in precision risk stratification of type 1 diabetes mellitus.

Type 1 diabetes (T1D) is a chronic metabolic disease mediated by autoimmunity. Its pathogenesis involves complex interactions between genetic susceptibility and environmental factors. Conventional T1D risk stratification primarily relies on genetic markers, islet autoantibodies, and glycemic indicators. Although these biomarkers remain indispensable in current clinical practice, they are often insufficient when used alone to accurately identify ultra-early high-risk individuals, predict disease progression rates, or support individualized preventive strategies. Consequently, more comprehensive molecular approaches are needed to improve precision risk stratification. In recent years, the rapid development of multi-omics technologies has provided new strategies for precise risk stratification of T1D. This narrative review critically evaluates how multi-omics integration strategies can improve precision risk stratification throughout the T1D disease continuum by integrating complementary molecular information from genomics, transcriptomics, proteomics, metabolomics, epigenomics, and the microbiome. Particular emphasis is placed on stage-specific biomarker discovery, multi-omics data integration frameworks, artificial intelligence-assisted prediction models, biomarker validation, and the opportunities and challenges associated with clinical translation. Current evidence suggests that integrated multi-omics approaches have the potential to improve risk prediction accuracy, distinguish heterogeneous disease trajectories, identify individuals at imminent risk of progression, and provide biologically informed targets for precision intervention. However, important challenges remain, including data harmonization, external validation, model interpretability, cost-effectiveness, and integration into routine clinical screening programs. Future research should prioritize prospective multicenter cohorts, standardized analytical pipelines, externally validated prediction models, and clinically interpretable multi-omics frameworks to facilitate the translation of precision risk stratification into routine T1D prevention and management.

Humans↗

A pan-cancer multi-omic SuperLearner for regulated cell death survival topologies.

INTRODUCTION: Regulated cell death (RCD) pathways influence tumor progression and immune modulation. We previously constructed a signature database mapping 25 RCD forms across seven multi-omic layers and 33 tumor types (CancerRCDShiny). Despite their ability to identify risk populations, translating these signatures into personalized clinical workflows requires a shift from cohort stratification to individualized risk mapping by modeling patient risk (survival topologies) to capture the non-linear dynamics of RCD signatures. METHODS: We engineered a pan-cancer multi-omic SuperLearner pipeline across 33 cancer types. Phase I performed zero-leakage harmonization and groupwise imputation to prevent cross-cohort amalgamation. Phase II deployed Elastic Net-regularized Cox regression as a CANARY diagnostic to map proportional hazards failures. Strata with a 35% missingness barrier entered Phase III, deploying a Quadripartite ensemble: Random Survival Forests, XGBoost, Survival-Boruta, and Multi-Task Logistic Regression, fused within an Elastic Net Multi-View Meta-Learner (MVL), with post-hoc TreeSHAP and LIME interpretability. RESULTS: The CANARY diagnostic demonstrated the structural invalidity of pan-cancer geometric proportional hazards. Across 96 admissible strata, Phase III executed algorithmic displacement: continuous multi-omic topologies suppressed static genomic mutations and copy number variations (85.7% vs. 0.0% apex retention). The MVL stabilized predictions against extreme variance; LIME surrogate validations (R 2&#x202f;<&#x202f;0.10) confirmed the systematic failure of linear interpretative proxies. N-dimensional TreeSHAP interaction mapping exposed synergistic and antagonistic rescue trajectories defining individualized Survival Topologies, which were invisible to additive models. The architecture was deployed as CancerRCDPredictor, a digital molecular tumor board with integrated LLM capabilities. The MVL SuperLearner achieved a median C-index of 0.749 (IQR: 0.722-0.836) across 96 modelable strata, with 95% bootstrap confidence intervals confirming precision (median width: 0.052) and permutation significance in 93.8% of strata (p&#x202f;<&#x202f;0.001). External CPTAC validation across ten cancer types demonstrated significant cross-cohort generalizability in clear cell renal carcinoma (KIRC; C-index 0.675, p&#x202f;=&#x202f;0.017) and modest performance across the remaining adequately powered cancers (median 0.582), underscoring the need for larger multi-institutional validation cohorts. CONCLUSION: This pan-cancer multi-omic SuperLearner bypasses linear topological failures, advancing beyond generalized stratification to establish a deterministically mapped architecture for predicting RCD-related survival topologies. Through the CancerRCDPredictor interface, multi-omic insights translate into individualized survival topology exploration, providing a foundation for future precision oncology validation.

SuperLearner↗

Concept of sample in OMICS technology.

Fundamental biological processes can now be studied by applying the full range of OMICS technologies (genomics, transcriptomics, proteomics, metabolomics, and beyond) to the same biological sample. Clearly, it would be desirable if the concept of sample were shared among these technologies, especially as up until the time a biological sample is prepared for use in a specific OMICS assay, its description is inherently technology independent. Sharing a common informatic representation would encourage data sharing (rather than data replication), thereby reducing redundant data capture and the potential for error. This would result in a significant degree of harmonization across different OMICS data standardization activities, a task that is critical if we are to integrate data from these different data sources. Here, we review the current concept of sample in OMICS technologies as it is being dealt with by different OMICS standardization initiatives and discuss the special role that the newly formed Genomic Standards Consortium (GSC) might have to play in this domain.

Animals↗

T-SMmOTE: tweaked synthetic majority minority oversampling technique for data scarcity issue in multi omics studies.

MOTIVATION: Multiomics data offer a rich data mine for modeling complex as well as day-to-day diseases, but their practical deployment is constrained by the limited sample availability. To this end, generating synthetic samples is a viable remedy. Extant schemes operating along this line, however, are mostly limited to augmenting the minority class in imbalanced datasets and often produce synthetic samples that lack sufficient diversity and fail to faithfully capture the underlying data distribution. As a result, the full potential of synthetic augmentation in multi-omics learning remains underexplored. The aim is to address the data scarcity problem in multi-omics domain. We propose a synthetic oversampling framework, which is dedicated to addressing overall data scarcity in multi-omics datasets and the lack of diversity in synthetic samples. Contrary to conventional methods that restrict augmentation to minority classes and rely on interpolation of two neighbors, our method generates diverse yet distribution-aligned synthetic samples by interpolating three neighbors and extends this augmentation paradigm to the majority class. The framework first balances the dataset by generating synthetic minority samples, and subsequently augments the balanced dataset by oversampling both majority and minority classes. RESULTS: Empirical evaluation on multi-omics data obtained from three heterogeneous health scenarios-inflammatory bowel disease, multi-organ dysfunction syndrome, and colorectal cancer-substantiates the utility of the proposed scheme in improving the predictive performance. The models trained on T-SMmOTE-augmented data achieve higher Matthews correlation coefficient values, along with improvedscores for both majority and minority classes. Notably, oversampling of the majority class improves the cognition of the minority class as well. We also explore the consistency of the class distributions between the original and augmented class-specific datasets. These findings confirm the capability of our scheme to learn from small, high-dimensional multi-omics datasets and highlight its potential for non-invasive disease detection. AVAILABILITY AND IMPLEMENTATION: https://github.com/payelu/TSMm.

Journal Article↗

IGCN: integrative graph convolution networks for patient level insights and biomarker discovery in multi-omics integration.

MOTIVATION: Developing computational tools for integrative analysis across multiple types of omics data has been of immense importance in cancer molecular biology and precision medicine research. While recent advancements have yielded integrative prediction solutions for multi-omics data, these methods lack a comprehensive and cohesive understanding of the rationale behind their specific predictions. To shed light on personalized medicine and unravel previously unknown characteristics within integrative analysis of multi-omics data, we introduce a novel integrative neural network approach for cancer molecular subtype and biomedical classification applications, named Integrative Graph Convolutional Networks (IGCN). RESULTS: To demonstrate the superiority of IGCN, we compare its performance with other state-of-the-art approaches across different cancer subtype and biomedical classification tasks. Our experimental results show that our proposed model outperforms the state-of-the-art and baseline methods. IGCN identifies which types of omics data receive more emphasis for each patient when predicting a specific class. Additionally, IGCN has the capability to pinpoint significant biomarkers from a range of omics data types. AVAILABILITY AND IMPLEMENTATION: The source code is available at https://github.com/bozdaglab/IGCN.

Humans↗

NOODAI: a webserver for network-oriented multi-omics data analysis and integration pipeline.

SUMMARY: Omics profiling has proven of great use for unbiased and comprehensive identification of key features that define biological phenotypes and underlie medical conditions. While each omics profile assists characterization of specific molecular components relevant for the studied phenotype, their joint evaluation can offer deeper insights into the overall mechanistic functioning of biological systems. Here, we introduce an approach where, starting from representative traits (e.g. differentially expressed elements) obtained for each omics profile, we construct and analyze joint interaction networks. The resulting networks rely on the existing knowledge of confident interactions among biological entities. We use these maps to identify and describe central elements, which connect multiple entities characteristic of the studied phenotypes and we leverage MONET network decomposition tool in order to highlight functionally connected network modules. In order to enable broad usage of this approach, we developed the NOODAI software platform, which enables integrative omics analysis through a user-friendly interface. The analysis outcomes are presented both as raw output tables as well as informative summary plots and written reports. Since the MONET tool enables the use of algorithms with strong performance in identifying disease-relevant modules, NOODAI software platform can be of a high value for analyzing clinical multi-omics datasets. AVAILABILITY AND IMPLEMENTATION: NOODAI is freely accessible at https://omics-oracle.com. Source code is available under GPL3 at: https://github.com/TotuTiberiu/NOODAI with the DOI: 10.5281/zenodo.17203984.

Software↗

OmicsPred as a centralised resource for genetic prediction of multi-omic traits.

Genetic prediction of multi-omic data has emerged as a cost-effective alternative to direct omics profiling, particularly useful for identifying molecular features associated with disease susceptibility. However, despite its popularity, multi-omic imputation models are fragmented across studies, hindering findability, accessibility, interoperability and re-use. To address this, we developed OmicsPred (https://www.omicspred.org), a centralised platform for the deposition and dissemination of genetic prediction models of multi-omic traits. OmicsPred unifies the most commonly used molecular imputation models (e.g. from PredictDB) and other published studies totalling 3,339,469 prediction models spanning transcriptomic, proteomic, and metabolomic traits (as of May 2026). Each model is accompanied by metadata describing score development and predictive performance, and distributed in formats compatible with popular analytic tools, such as PGS Catalog Calculator and MetaXcan. To demonstrate the utility of the resource for systematic target discovery, we perform a multi-omic phenome-wide association analysis in Million Veterans Program data.

Journal Article↗

Bridging genotype, phenotype, and clinical insight: the role of multi-omics in cardiovascular disease.

INTRODUCTION: It is increasingly evident that the multifactorial nature of cardiovascular disease requires the combination of different omics approaches for improving our mechanistic understanding, identifying novel drug targets, and developing accurate diagnostic, predictive, and prognostic biomarker panels. AREAS COVERED: We review the current state and the potential of multi-omics in cardiovascular disease, with a specific focus on plasma-, spatial-, and single-cell approaches. We discuss lipidomics as a genotype&#x2011;to&#x2011;phenotype bridge, the utility of remote longitudinal monitoring via microsampling/dried blood spots, and emerging clinical&#x2011;trial integrations of multi-omics approaches. We outline critical gaps in standardization and how to overcome these, pre&#x2011;analytical challenges and constraints that are often neglected, and data&#x2011;integration methods spanning from canonical correlation analysis to modern machine learning approaches. EXPERT OPINION: Multi&#x2011;omics can shape cardiovascular care by identifying drug targets in diseased tissue and by yielding small, usable biomarker panels.

Humans↗

Multi&#x2011;omics approaches to decipher the molecular mechanisms of exercise&#x2011;mediated bone protection: From mechanistic insights to personalized exercise prescription (Review).

The global burden of bone metabolic disorders necessitates a shift from generic exercise recommendations toward personalized prescription strategies. Exercise confers skeletal protection through mechanotransduction, yet the underlying molecular networks remain incompletely understood. Multi&#x2011;omics technologies, including transcriptomics, proteomics, metabolomics and single&#x2011;cell spatial approaches, have revolutionized the capacity to decode exercise&#x2011;mediated bone adaptation at the systems level. The present review synthesizes current single&#x2011;omics landscapes and integrative multi&#x2011;omics analyses that elucidate the core regulatory networks, mechanobiological coupling mechanisms and multiorgan crosstalk that are implicated in the bone response to mechanical loading. Translational applications across clinical scenarios such as osteoporosis, osteoarthritis and disuse bone loss are evaluated, and the technical, analytical and translational challenges limiting clinical implementation are addressed. Finally, the present review provides a framework for translating multi&#x2011;omics molecular signatures into personalized exercise prescriptions for optimized skeletal health.

Humans↗

Livestock Multi-Omics Integration: A Systematic Framework From Statistical Association to Causal Interpretation.

Livestock multi-omics integration is key to unraveling complex trait regulation, yet systematic, livestock-specific strategies remain scarce. This review traces the progression from single-omics accumulation to multi-dimensional integration, highlighting how large-scale genomic, epigenomic, and transcriptomic projects lay the foundation for functional dissection. We identify core impediments: extreme species diversity, marked data heterogeneity, limited sample sizes, and a pervasive reduction of multi-omics data to simplistic differential screens, resulting in low translational efficiency. We critically appraise four common pitfalls-overinterpreting correlation as causation, relegating proteomics to corroborating transcriptomics, incomplete microbiome-host integration lacking environmental context, and systematic neglect of metabolic fluxomics-and show how exposomics and fluxomics add necessary causal and dynamic dimensions. To address these, we propose a livestock-adapted three-tier analytical framework: (1) statistical association of cross-omics covariation patterns; (2) machine learning-driven feature mining and integrative modeling; and (3) causal interpretation encompassing Mendelian randomization, prior-knowledge-guided network inference, and physical causal evidence via fluxomics and metabolic control analysis. We further discuss how multimodal sequencing (single-cell, spatial, temporal) and generative AI can fundamentally mitigate heterogeneity and strengthen causal evidence. Finally, we outline future priorities in database standardization, livestock-specific benchmarking, and translational pipelines, charting a path from correlation-centric reporting to mechanistic causality and precision breeding.

Animals↗

Omics in hereditary optic neuropathies: A systematic review of clinical studies with an integrated point of view.

Hereditary optic neuropathies are characterized by bilateral visual loss due to the degeneration of retinal ganglion cells, resulting in optic nerve degeneration and atrophy. Although the genetic origin of the main isolated and syndromic hereditary optic neuropathies has been characterized, the clinical phenotypes exhibit significant and poorly understood variability in both penetrance and expressivity. Additionally, the genetic and environmental factors that influence the onset of these optic neuropathies remain poorly understood, with limited biomarkers to predict disease progression or as readouts for therapeutic trials. Data-driven omics strategies allow deep phenotyping to improve our understanding of pathophysiological mechanisms and to search for new biomarkers and therapeutic targets. We explore whether the omics strategies applied to patients with hereditary optic neuropathies have provided such new insights. MEDLINE, Web of Science and EMBASE databases were screened for studies with terms relating to hereditary optic neuropathies, transcriptomics, epigenomics, proteomics, metabolomics and lipidomics in clinical studies exploring patients' samples. Out of 1244 references identified, 22 articles were included after double-masked data curation. These articles focused only on the 3 main forms of hereditary optic neuropathies, namely, OPA1-related dominant optic atrophy (n&#x202f;=&#x202f;4), Leber hereditary optic neuropathy (n&#x202f;=&#x202f;13), and Wolfram syndrome (n&#x202f;=&#x202f;5). While the methodological designs and results of these studies were highly heterogeneous, they revealed molecular alterations that we have attempted to discuss at the integrated multi-omics level. This data integration highlighted several common pathophysiological mechanisms such as energetic impairment, endoplasmic reticulum stress, proteotoxic and oxidative stresses, lipid remodeling and altered amino acid and purine metabolisms, while suggesting potential new biomarkers and therapeutic targets. These findings underscore the potential of integrated multi-omics approaches to deepen our understanding of the phenotypic complexity of hereditary optic neuropathies and to support the development of innovative diagnostic and therapeutic strategies.

Humans↗

A cross-species multi-omics analyze uncovers conserved molecular mechanisms underlying age-related erectile dysfunction.

BACKGROUND: The urgent need for new treatments is driven by the challenging clinical situation of age-related erectile dysfunction (ARED). AIM: To clarify the conserved molecular mechanisms of ARED across species using multi-omics. METHODS: Rat and mouse models with ARED were developed to facilitate the extraction of mRNA and proteins from the corpus cavernosum for high-throughput sequencing. Bioinformatics techniques were employed to analyze differentially expressed genes and to conduct analyses using the Kyoto Encyclopedia of Genes and Genomes, Gene Ontology, and protein-protein interaction networks. Verification of the results was carried out using immunofluorescence, hematoxylin-eosin staining, and Masson staining. OUTCOMES: The multi-omics profiles of ARED rats and mice were analyzed and validated across species. RESULTS: In both species, Kyoto Encyclopedia of Genes and Genomes and Gene Ontology analyses of transcriptomic and proteomic data revealed that differentially expressed genes were predominantly enriched in pathways associated with alterations in extracellular matrix composition, downregulation of mitochondrial activity, and disruption of protein homeostasis. Immunofluorescence analysis demonstrated an upregulation of reactive oxygen species expression, coupled with a downregulation of Aldh18a1, collagen, and collagen I expression in the corpus cavernosum of mice and rats with ARED. CLINICAL IMPLICATIONS: To offer a novel approach for enhancing the erectile function in patients with ARED. STRENGTHS AND LIMITATIONS: The primary strength of this study lies in its utilization of cross-species multi-omics sequencing, which has elucidated the conserved molecular mechanisms underlying ARED. However, a significant limitation is the absence of subsequent validation in patients with ARED. CONCLUSIONS: Cross-species multi-omics comparisons present a potentially innovative approach for elucidating the underlying mechanisms and identifying preventive and therapeutic targets for ARED.

aging↗

Designing smart spatial omics experiments with S2Omics.

Spatial omics technologies have transformed biomedical research by enabling high-resolution molecular profiling while preserving the native tissue architecture. These advances provide unprecedented insights into tissue structure and function. However, the high cost and time-intensive nature of spatial omics experiments necessitate careful experimental design, particularly in selecting regions of interest (ROIs) from large tissue sections. Currently, ROI selection is performed manually, which introduces subjectivity, inconsistency, and a lack of reproducibility. Previous studies have shown strong correlations between spatial molecular patterns and histological features, suggesting that readily available and cost-effective histology images can be leveraged to guide spatial omics experiments. Here, we present S2Omics, an end-to-end workflow that automatically selects ROIs from histology images with the goal of maximizing molecular information content in the ROIs. Through comprehensive evaluations across multiple spatial omics platforms and tissue types, we demonstrate that S2Omics enables systematic and reproducible ROI selection and enhances the robustness and impact of downstream biological discovery.

digital pathology↗

MetaServe: a lightweight, metadata-aware governance and delivery layer for pre-publication research omics data.

BACKGROUND: Institutional research teams and core facilities routinely manage pre-publication omics datasets that span heterogeneous file types, nested project structures, and multiple downstream uses. Public repositories mainly support post-publication dissemination, while workflow systems and enterprise data platforms do not directly provide a lightweight governance and delivery layer for internal research assets. RESULTS: We present MetaServe, an open-source governance and delivery layer for pre-publication research assets in institutional multi-omics settings. MetaServe registers and delivers heterogeneous assets, including sequencing files, processed matrices, imaging data, analysis-ready objects, tabular files, and documents, without requiring repository-grade standardization. Its metadata-aware design combines file-type recognition, partial automatic extraction for selected formats, manually supplied project and biological annotations, and indexed faceted retrieval. MetaServe supports authenticated web download, viewer-oriented handoff for compatible services such as cellxgene, and path-manifest export for downstream workflows under shared-storage assumptions. The current implementation combines role-based controls, explicit file-level sharing, path-constrained delivery, and operational traceability to support controlled institutional access. MetaServe has been deployed at the Chinese Institutes for Medical Research (CIMR) as part of an institutional multi-omics data-management system. CONCLUSIONS: MetaServe provides a practical layer between institutional storage and downstream analytical platforms for pre-publication research data. Its contribution is the integration of lightweight metadata-aware registration, permission-aware retrieval, and controlled delivery for heterogeneous institutional omics assets. Rather than replacing workflow engines, public repositories, or enterprise-scale research data platforms, MetaServe offers a deployable governance layer for core facilities and collaborative teams that need structured discovery and traceable delivery before public deposition or manuscript release.

Metadata↗

HINN: Hierarchical Input Neural Network identifies multi-omics biomarker for cognitive decline.

Understanding complex diseases requires models that can integrate diverse layers of biological data while yielding insights that are biologically interpretable. Although multi-omics integration with machine learning (ML) has advanced disease prediction and biomarker discovery, most existing approaches overlook the hierarchical and regulatory relationships that connect these molecular layers. Here, we present the Hierarchical Input Neural Network (HINN), a deep learning framework that incorporates known cross-omics relationships directly into its architecture, capturing the flow of information from genomics to epigenomics, transcriptomics, and downstream biological processes. By embedding these relationships, HINN improves both predictive performance and biological interpretability. We applied HINN to blood-derived multi-omics data from individuals with Alzheimer's disease or mild cognitive impairment to predict cognitive scores from standardized assessments. HINN outperformed both baseline and state-of-the-art models and pinpointed multi-omics biomarkers-including SNPs and promoter-region CpG sites in ATP6V1C1 and RCHY1 -that were significantly correlated with plasma p-Tau181 levels. These features map to biologically relevant processes with potential implications for cognitive decline. Our findings demonstrate how combining deep learning with biological knowledge can uncover interpretable, blood-based biomarkers for cognitive decline due to complex diseases such as Alzheimer's. All code and data are openly available at https://github.com/bozdaglab/HINN.

Alzheimer&#x2019;s disease↗