Clustering and variable selection evaluation of 13 unsupervised methods for multi-omics data integration

Benchmarking multi-omics integrative clustering methods for subtype identification in colorectal cancer

BACKGROUND AND OBJECTIVE

Colorectal cancer (CRC) represents a heterogeneous malignancy that has concerned global burden of incidence and mortality. The traditional tumor-node-metastasis staging system has exhibited certain limitations. With the advancement of omics technologies, researchers are directing their focus on developing a more precise multi-omics molecular classification. Therefore, the utilization of unsupervised multi-omics integrative clustering methods in CRC, advocating for the establishment of a comprehensive benchmark with practical guidelines.

METHODS

In this study, we obtained CRC multi-omics data, encompassing DNA methylation, gene expression, and protein expression from the cancer genome atlas (TCGA)database. We then generated interrelated CRC multi-omics data with various structures based on realistic multi-omics correlations, and performed a comprehensive evaluation of eight representative methods categorized as early integration, intermediate integration, and late integration using complementary benchmarks for subtype classification accuracy. Lastly, we employed these methods to integrate real-world CRC multi-omics data, survival and differential analysis were used to highlight differences among newly identified multi-omics subtypes.

RESULTS

Through in-depth comparisons, we observed that similarity network fusion (SNF) exhibited exceptional performance in integrating multi-omics data derived from simulations. Additionally, SNF effectively distinguished CRC patients into five subgroups with the highest classification accuracy. Moreover, we found significant survival differences and molecular distinctions among SNF subtypes.

CONCLUSIONS

The findings consistently demonstrate that SNF outperforms other methods in CRC multi-omics integrative clustering. The significant survival differences and molecular distinctions among SNF subtypes provide novel insights into the multi-omics perspective on CRC heterogeneity with potential clinical treatment.

Effective data visualization strategies in untargeted metabolomics

Covering: 2014 to 2023 for metabolomics, 2002 to 2023 for information visualizationLC-MS/MS-based untargeted metabolomics is a rapidly developing research field spawning increasing numbers of computational metabolomics tools assisting researchers with their complex data processing, analysis, and interpretation tasks. In this article, we review the entire untargeted metabolomics workflow from the perspective of information visualization, visual analytics and visual data integration. Data visualization is a crucial step at every stage of the metabolomics workflow, where it provides core components of data inspection, evaluation, and sharing capabilities. However, due to the large number of available data analysis tools and corresponding visualization components, it is hard for both users and developers to get an overview of what is already available and which tools are suitable for their analysis. In addition, there is little cross-pollination between the fields of data visualization and metabolomics, leaving visual tools to be designed in a secondary and mostly ad hoc fashion. With this review, we aim to bridge the gap between the fields of untargeted metabolomics and data visualization. First, we introduce data visualization to the untargeted metabolomics field as a topic worthy of its own dedicated research, and provide a primer on cutting-edge visualization research into data visualization for both researchers as well as developers active in metabolomics. We extend this primer with a discussion of best practices for data visualization as they have emerged from data visualization studies. Second, we provide a practical roadmap to the visual tool landscape and its use within the untargeted metabolomics field. Here, for several computational analysis stages within the untargeted metabolomics workflow, we provide an overview of commonly used visual strategies with practical examples. In this context, we will also outline promising areas for further research and development. We end the review with a set of recommendations for developers and users on how to make the best use of visualizations for more effective and transparent communication of results.

Machine learning-based clustering identifies obesity subgroups with differential multi-omics profiles and metabolic patterns

OBJECTIVE

Individuals living with obesity are differentially susceptible to cardiometabolic diseases. We hypothesized that an integrative multi-omics approach might improve identification of subgroups of individuals with obesity who have distinct cardiometabolic disease patterns.

METHODS

We performed machine learning-based, integrative unsupervised clustering to identify proteomics- and metabolomics-defined subpopulations of individuals living with obesity (BMI ≥ 30 kg/m2), leveraging data from 243 individuals in the Multi-Ethnic Study of Atherosclerosis (MESA) cohort. Omics that contributed to the observed clusters were functionally characterized. We performed multivariate regression to assess whether the individuals in each cluster demonstrated differential patterns of cardiometabolic traits.

RESULTS

We identified two distinct clusters (iCluster1 and 2). iCluster2 had significantly higher average BMI values, fasting blood glucose, and inflammation. iCluster1 was associated with higher levels of total cholesterol and high-density lipoprotein cholesterol. Pathways mediating cell growth, lipogenesis, and energy expenditures were positively associated with iCluster1. Inflammatory response and insulin resistance pathways were positively associated with iCluster2.

CONCLUSIONS

Although the two identified clusters may represent progressive obesity-related pathologic processes measured at different stages, other mechanisms in combination could also underpin the identified clusters given no significant age difference between the comparative groups. For instance, clusters may reflect differences in dietary/behavioral patterns or differential rates of metabolic damage.

A comprehensive review of machine learning techniques for multi-omics data integration: challenges and applications in precision oncology

Multi-omics data play a crucial role in precision medicine, mainly to understand the diverse biological interaction between different omics. Machine learning approaches have been extensively employed in this context over the years. This review aims to comprehensively summarize and categorize these advancements, focusing on the integration of multi-omics data, which includes genomics, transcriptomics, proteomics and metabolomics, alongside clinical data. We discuss various machine learning techniques and computational methodologies used for integrating distinct omics datasets and provide valuable insights into their application. The review emphasizes both the challenges and opportunities present in multi-omics data integration, precision medicine and patient stratification, offering practical recommendations for method selection in various scenarios. Recent advances in deep learning and network-based approaches are also explored, highlighting their potential to harmonize diverse biological information layers. Additionally, we present a roadmap for the integration of multi-omics data in precision oncology, outlining the advantages, challenges and implementation difficulties. Hence this review offers a thorough overview of current literature, providing researchers with insights into machine learning techniques for patient stratification, particularly in precision oncology. Contact:  anirban@klyuniv.ac.in.

Molecular precision medicine: Multi-omics-based stratification model for acute myeloid leukemia

Acute myeloid leukemia (AML), as the most common malignancy of the hematopoietic system, poses challenges in treatment efficacy, relapse, and drug resistance. In this study, we have utilized 151 RNA sequencing datasets, 194 DNA methylation datasets, and 200 somatic mutation datasets from the AML cohort in the TCGA database to develop a multi-omics stratification model. This model enables comparison of prognosis, clinical features, gene mutations, immune microenvironment and drug sensitivity across subgroups. External validation datasets have been sourced from the GEO database, which includes 562 mRNA datasets and 136 miRNA datasets from 984 adult AML patients. Through multi-omics-based stratification model, we classified 126 AML patients into 4 clusters (CS). CS4 had the best prognosis, with the youngest age, highest M3 subtype proportion, fewest copy number alterations, and common mutations in WT1, FLT3, and KIT genes. It showed sensitivity to HDAC inhibitors and BCL-2 inhibitors. Both the M3 subtype and CS4 were identified as independent protective factors for survival. Conversely, CS3 had the worst prognosis due to older age, high copy number alterations, and frequent mutations in RUNX1, DNMT3A, and TP53 genes. Additionally, it showed higher proportions of cytotoxic cells and Tregs, suggesting potential sensitivity to mTOR inhibitors. CS1 had a better prognosis than CS2, with more copy number alterations, while CS2 had higher monocyte proportions. CS1 showed good sensitivity to cytarabine, while CS2 was sensitive to RXR agonists. Both CS1 and CS2, which predominantly featured mutations in FLT3, NPM1, and DNMT3A genes, benefited from FLT3 inhibitors. Using the Kappa test, our stratification model underwent robust validation in the miRNA and mRNA external validation datasets. With advancements in sequencing technology and machine learning algorithms, AML is poised to transition towards multi-omics precision medicine in the future. We aspire for our study to offer new perspectives on multi-drug combination clinical trials and multi-targeted precision medicine for AML.

An extension of latent unknown clustering integrating multi-omics data (LUCID) incorporating incomplete omics data

Motivation

Latent unknown clustering integrating multi-omics data is a novel statistical model designed for multi-omics data analysis. It integrates omics data with exposures and an outcome through a latent cluster, elucidating how exposures influence processes reflected in multi-omics measurements, ultimately affecting an outcome. A significant challenge in multi-omics analysis is the issue of list-wise missingness. To address this, we extend the model to incorporate list-wise missingness within an integrated imputation framework, which can also handle sporadic missingness when necessary.

Results

Simulation studies demonstrate that our integrated imputation approach produces consistent and less biased estimates, closely reflecting true underlying values. We applied this model to data from the ISGlobal/ATHLETE "Exposome Data Challenge Event" to explore the association between maternal exposure to hexachlorobenzene and childhood body mass index by integrating incomplete proteomics data from 1301 children. The model successfully estimated proteomics profiles for two clusters representing higher and lower body mass index, characterizing the potential profiles linking prenatal hexachlorobenzene levels and childhood body mass index.

Availability and implementation

The proposed methods have been implemented in the R package LUCIDus. The source code is available at https://github.com/USCbiostats/LUCIDus.

COPS: A novel platform for multi-omic disease subtype discovery via robust multi-objective evaluation of clustering algorithms

Recent research on multi-view clustering algorithms for complex disease subtyping often overlooks aspects like clustering stability and critical assessment of prognostic relevance. Furthermore, current frameworks do not allow for a comparison between data-driven and pathway-driven clustering, highlighting a significant gap in the methodology. We present the COPS R-package, tailored for robust evaluation of single and multi-omics clustering results. COPS features advanced methods, including similarity networks, kernel-based approaches, dimensionality reduction, and pathway knowledge integration. Some of these methods are not accessible through R, and some correspond to new approaches proposed with COPS. Our framework was rigorously applied to multi-omics data across seven cancer types, including breast, prostate, and lung, utilizing mRNA, CNV, miRNA, and DNA methylation data. Unlike previous studies, our approach contrasts data- and knowledge-driven multi-view clustering methods and incorporates cross-fold validation for robustness. Clustering outcomes were assessed using the ARI score, survival analysis via Cox regression models including relevant covariates, and the stability of the results. While survival analysis and gold-standard agreement are standard metrics, they vary considerably across methods and datasets. Therefore, it is essential to assess multi-view clustering methods using multiple criteria, from cluster stability to prognostic relevance, and to provide ways of comparing these metrics simultaneously to select the optimal approach for disease subtype discovery in novel datasets. Emphasizing multi-objective evaluation, we applied the Pareto efficiency concept to gauge the equilibrium of evaluation metrics in each cancer case-study. Affinity Network Fusion, Integrative Non-negative Matrix Factorization, and Multiple Kernel K-Means with linear or Pathway Induced Kernels were the most stable and effective in discerning groups with significantly different survival outcomes in several case studies.

Comprehensive evaluation of disulfidptosis in intestinal immunity and biologic therapy response in Ulcerative Colitis

Objective

Ulcerative Colitis (UC) manifests as a chronic inflammatory condition of the intestines, marked by ongoing immune system dysregulation. Disulfidptosis, a newly identified cell death mechanism, is intimately linked to the onset and advancement of inflammation. However, the role of disulfidptosis in UC remains unclear.

Methods

We screened differentially expressed genes (DEGs) associated with disulfidptosis in multiple UC datasets, narrowed down the target gene number using lasso regression, and conducted immune infiltration analysis and constructed a clinical diagnostic model. Additionally, we explored the association between disulfidptosis-related key genes and disease remission in UC patients receiving biologic therapy. Finally, we confirmed the expression of key genes in FHC cells and UC tissue samples.

Results

In the differential analysis, we identified 20 DEGs associated with disulfidptosis. Immune infiltration results revealed that five genes (PDLIM1, SLC7A11, MYH10, NUBPL, OXSM) exhibited strong correlations with immune cells and pathways. Using GO, KEGG and WGCNA analyses, we discovered that gene modules highly correlated with disulfidptosis-related gene expression were significantly enriched in inflammation-related pathways. Additionally, we developed a nomogram based on these five immune-related disulfidptosis genes for UC diagnosis, showing robust diagnostic capability and clinical efficacy. Kaplan-Meier survival analysis revealed a significant link between changes in the expression levels of these cell genes and disease remission in UC patients receiving biologic therapy. In line with previous studies, similar expression changes of the target gene were seen in both UC cell models and tissue samples.

Conclusions

This study utilized bioinformatic analysis and machine learning to identify and analyze features associated with disulfidptosis in multiple UC datasets. This enhances our comprehension of the role disulfidptosis plays in intestinal immunity and inflammation in UC, providing new perspectives for developing innovative treatments for UC.

Unveiling divergent treatment prognoses in IDHwt-GBM subtypes through multiomics clustering: a swift dual MRI-mRNA model for precise subtype prediction

BACKGROUND

IDH1-wildtype glioblastoma multiforme (IDHwt-GBM) is a highly heterogeneous and aggressive brain tumour characterised by a dismal prognosis and significant challenges in accurately predicting patient outcomes. To address these issues and personalise treatment approaches, we aimed to develop and validate robust multiomics molecular subtypes of IDHwt-GBM. Through this, we sought to uncover the distinct molecular signatures underlying these subtypes, paving the way for improved diagnosis and targeted therapy for this challenging disease.

METHODS

To identify stable molecular subtypes among 184 IDHwt-GBM patients from TCGA, we used the consensus clustering method to consolidate the results from ten advanced multiomics clustering approaches based on mRNA, lncRNA, and mutation data. We developed subtype prediction models using the PAM and machine learning algorithms based on mRNA and MRI data for enhanced clinical utility. These models were validated in five independent datasets, and an online interactive system was created. We conducted a comprehensive assessment of the clinical impact, drug treatment response, and molecular associations of the IDHwt-GBM subtypes.

RESULTS

In the TCGA cohort, two molecular subtypes, class 1 and class 2, were identified through multiomics clustering of IDHwt-GBM patients. There was a significant difference in survival between Class 1 and Class 2 patients, with a hazard ratio (HR) of 1.68 [1.15-2.47]. This difference was validated in other datasets (CGGA: HR = 1.75[1.04, 2.94]; CPTAC: HR = 1.79[1.09-2.91]; GALSS: HR = 1.66[1.09-2.54]; UCSF: HR = 1.33[1.00-1.77]; UPENN HR = 1.29[1.04-1.58]). Additionally, class 2 was more sensitive to treatment with radiotherapy combined with temozolomide, and this sensitivity was validated in the GLASS cohort. Correspondingly, class 2 and class 1 exhibited significant differences in mutation patterns, enriched pathways, programmed cell death (PCD), and the tumour immune microenvironment. Class 2 had more mutation signatures associated with defective DNA mismatch repair (P = 0.0021). Enriched pathways of differentially expressed genes in class 1 and class 2 (P-adjust < 0.05) were mainly related to ferroptosis, the PD-1 checkpoint pathway, the JAK-STAT signalling pathway, and other programmed cell death and immune-related pathways. The different cell death modes and immune microenvironments were validated across multiple datasets. Finally, our developed survival prediction model, which integrates molecular subtypes, age, and sex, demonstrated clinical benefits based on the decision curve in the test set. We deployed the molecular subtyping prediction model and survival prediction model online, allowing interactive use and facilitating user convenience.

CONCLUSIONS

Molecular subtypes were identified and verified through multiomics clustering in IDHwt-GBM patients. These subtypes are linked to specific mutation patterns, the immune microenvironment, prognoses, and treatment responses.

Subgroup identification-based model selection to improve the predictive performance of individualized dosing

Currently, model-informed precision dosing uses one population pharmacokinetic model that best fits the target population. We aimed to develop a subgroup identification-based model selection approach to improve the predictive performance of individualized dosing, using vancomycin in neonates/infants as a test case. Data from neonates/infants with at least one vancomycin concentration was randomly divided into training and test dataset. Population predictions from published vancomycin population pharmacokinetic models were calculated. The single best-performing model based on various performance metrics, including median absolute percentage error (APE) and percentage of predictions within 20% (P20) or 60% (P60) of measurement, were determined. Clustering based on median APEs or clinical and demographic characteristics and model selection by genetic algorithm was used to group neonates/infants according to their best-performing model. Subsequently, classification trees to predict the best-performing model using clinical and demographic characteristics were developed. A total of 208 vancomycin treatment episodes in training and 88 in test dataset was included. Of 30 identified models from the literature, the single best-performing model for training dataset had P20 26.2-42.6% in test dataset. The best-performing clustering approach based on median APEs or clinical and demographic characteristics and model selection by genetic algorithm had P20 44.1-45.5% in test dataset, whereas P60 was comparable. Our proof-of-concept study shows that the prediction of the best-performing model for each patient according to the proposed model selection approaches has the potential to improve the predictive performance of model-informed precision dosing compared with the single best-performing model approach.

Comparative analysis of integrative classification methods for multi-omics data

Recent advances in sequencing, mass spectrometry, and cytometry technologies have enabled researchers to collect multiple 'omics data types from a single sample. These large datasets have led to a growing consensus that a holistic approach is needed to identify new candidate biomarkers and unveil mechanisms underlying disease etiology, a key to precision medicine. While many reviews and benchmarks have been conducted on unsupervised approaches, their supervised counterparts have received less attention in the literature and no gold standard has emerged yet. In this work, we present a thorough comparison of a selection of six methods, representative of the main families of intermediate integrative approaches (matrix factorization, multiple kernel methods, ensemble learning, and graph-based methods). As non-integrative control, random forest was performed on concatenated and separated data types. Methods were evaluated for classification performance on both simulated and real-world datasets, the latter being carefully selected to cover different medical applications (infectious diseases, oncology, and vaccines) and data modalities. A total of 15 simulation scenarios were designed from the real-world datasets to explore a large and realistic parameter space (e.g. sample size, dimensionality, class imbalance, effect size). On real data, the method comparison showed that integrative approaches performed better or equally well than their non-integrative counterpart. By contrast, DIABLO and the four random forest alternatives outperform the others across the majority of simulation scenarios. The strengths and limitations of these methods are discussed in detail as well as guidelines for future applications.

Challenges and best practices in omics benchmarking

Technological advances enabling massively parallel measurement of biological features - such as microarrays, high-throughput sequencing and mass spectrometry - have ushered in the omics era, now in its third decade. The resulting complex landscape of analytical methods has naturally fostered the growth of an omics benchmarking industry. Benchmarking refers to the process of objectively comparing and evaluating the performance of different computational or analytical techniques when processing and analysing large-scale biological data sets, such as transcriptomics, proteomics and metabolomics. With thousands of omics benchmarking studies published over the past 25 years, the field has matured to the point where the foundations of benchmarking have been established and well described. However, generating meaningful benchmarking data and properly evaluating performance in this complex domain remains challenging. In this Review, we highlight some common oversights and pitfalls in omics benchmarking. We also establish a methodology to bring the issues that can be addressed into focus and to be transparent about those that cannot: this takes the form of a spreadsheet template of guidelines for comprehensive reporting, intended to accompany publications. In addition, a survey of recent developments in benchmarking is provided as well as specific guidance for commonly encountered difficulties.

Multiomics data integration, limitations, and prospects to reveal the metabolic activity of the coral holobiont

Since their radiation in the Middle Triassic period ∼240 million years ago, stony corals have survived past climate fluctuations and five mass extinctions. Their long-term survival underscores the inherent resilience of corals, particularly when considering the nutrient-poor marine environments in which they have thrived. However, coral bleaching has emerged as a global threat to coral survival, requiring rapid advancements in coral research to understand holobiont stress responses and allow for interventions before extensive bleaching occurs. This review encompasses the potential, as well as the limits, of multiomics data applications when applied to the coral holobiont. Synopses for how different omics tools have been applied to date and their current restrictions are discussed, in addition to ways these restrictions may be overcome, such as recruiting new technology to studies, utilizing novel bioinformatics approaches, and generally integrating omics data. Lastly, this review presents considerations for the design of holobiont multiomics studies to support lab-to-field advancements of coral stress marker monitoring systems. Although much of the bleaching mechanism has eluded investigation to date, multiomic studies have already produced key findings regarding the holobiont's stress response, and have the potential to advance the field further.

A Generalized Higher-order Correlation Analysis Framework for Multi-Omics Network Inference

Multiple -omics (genomics, proteomics, etc.) profiles are commonly generated to gain insight into a disease or physiological system. Constructing multi-omics networks with respect to the trait(s) of interest provides an opportunity to understand relationships between molecular features but integration is challenging due to multiple data sets with high dimensionality. One approach is to use canonical correlation to integrate one or two omics types and a single trait of interest. However, these types of methods may be limited due to (1) not accounting for higher-order correlations existing among features, (2) computational inefficiency when extending to more than two omics data when using a penalty term-based sparsity method, and (3) lack of flexibility for focusing on specific correlations (e.g., omics-to-phenotype correlation versus omics-to-omics correlations). In this work, we have developed a novel multi-omics network analysis pipeline called Sparse Generalized Tensor Canonical Correlation Analysis Network Inference (SGTCCA-Net) that can effectively overcome these limitations. We also introduce an implementation to improve the summarization of networks for downstream analyses. Simulation and real-data experiments demonstrate the effectiveness of our novel method for inferring omics networks and features of interest.

Automated calibration of consensus weighted distance-based clustering approaches using sharp

MOTIVATION

In consensus clustering, a clustering algorithm is used in combination with a subsampling procedure to detect stable clusters. Previous studies on both simulated and real data suggest that consensus clustering outperforms native algorithms.

RESULTS

We extend here consensus clustering to allow for attribute weighting in the calculation of pairwise distances using existing regularized approaches. We propose a procedure for the calibration of the number of clusters (and regularization parameter) by maximizing the sharp score, a novel stability score calculated directly from consensus clustering outputs, making it extremely computationally competitive. Our simulation study shows better clustering performances of (i) approaches calibrated by maximizing the sharp score compared to existing calibration scores and (ii) weighted compared to unweighted approaches in the presence of features that do not contribute to cluster definition. Application on real gene expression data measured in lung tissue reveals clear clusters corresponding to different lung cancer subtypes.

AVAILABILITY AND IMPLEMENTATION

The R package sharp (version ≥1.4.3) is available on CRAN at https://CRAN.R-project.org/package=sharp.

Applications of multi-omics analysis in human diseases

Multi-omics usually refers to the crossover application of multiple high-throughput screening technologies represented by genomics, transcriptomics, single-cell transcriptomics, proteomics and metabolomics, spatial transcriptomics, and so on, which play a great role in promoting the study of human diseases. Most of the current reviews focus on describing the development of multi-omics technologies, data integration, and application to a particular disease; however, few of them provide a comprehensive and systematic introduction of multi-omics. This review outlines the existing technical categories of multi-omics, cautions for experimental design, focuses on the integrated analysis methods of multi-omics, especially the approach of machine learning and deep learning in multi-omics data integration and the corresponding tools, and the application of multi-omics in medical researches (e.g., cancer, neurodegenerative diseases, aging, and drug target discovery) as well as the corresponding open-source analysis tools and databases, and finally, discusses the challenges and future directions of multi-omics integration and application in precision medicine. With the development of high-throughput technologies and data integration algorithms, as important directions of multi-omics for future disease research, single-cell multi-omics and spatial multi-omics also provided a detailed introduction. This review will provide important guidance for researchers, especially who are just entering into multi-omics medical research.

Integrative multi-omics and systems bioinformatics in translational neuroscience: A data mining perspective

Bioinformatic analysis of large and complex omics datasets has become increasingly useful in modern day biology by providing a great depth of information, with its application to neuroscience termed neuroinformatics. Data mining of omics datasets has enabled the generation of new hypotheses based on differentially regulated biological molecules associated with disease mechanisms, which can be tested experimentally for improved diagnostic and therapeutic targeting of neurodegenerative diseases. Importantly, integrating multi-omics data using a systems bioinformatics approach will advance the understanding of the layered and interactive network of biological regulation that exchanges systemic knowledge to facilitate the development of a comprehensive human brain profile. In this review, we first summarize data mining studies utilizing datasets from the individual type of omics analysis, including epigenetics/epigenomics, transcriptomics, proteomics, metabolomics, lipidomics, and spatial omics, pertaining to Alzheimer's disease, Parkinson's disease, and multiple sclerosis. We then discuss multi-omics integration approaches, including independent biological integration and unsupervised integration methods, for more intuitive and informative interpretation of the biological data obtained across different omics layers. We further assess studies that integrate multi-omics in data mining which provide convoluted biological insights and offer proof-of-concept proposition towards systems bioinformatics in the reconstruction of brain networks. Finally, we recommend a combination of high dimensional bioinformatics analysis with experimental validation to achieve translational neuroscience applications including biomarker discovery, therapeutic development, and elucidation of disease mechanisms. We conclude by providing future perspectives and opportunities in applying integrative multi-omics and systems bioinformatics to achieve precision phenotyping of neurodegenerative diseases and towards personalized medicine.

Comparison of cancer subtype identification methods combined with feature selection methods in omics data analysis

BACKGROUND

Cancer subtype identification is important for the early diagnosis of cancer and the provision of adequate treatment. Prior to identifying the subtype of cancer in a patient, feature selection is also crucial for reducing the dimensionality of the data by detecting genes that contain important information about the cancer subtype. Numerous cancer subtyping methods have been developed, and their performance has been compared. However, combinations of feature selection and subtype identification methods have rarely been considered. This study aimed to identify the best combination of variable selection and subtype identification methods in single omics data analysis.

RESULTS

Combinations of six filter-based methods and six unsupervised subtype identification methods were investigated using The Cancer Genome Atlas (TCGA) datasets for four cancers. The number of features selected varied, and several evaluation metrics were used. Although no single combination was found to have a distinctively good performance, Consensus Clustering (CC) and Neighborhood-Based Multi-omics Clustering (NEMO) used with variance-based feature selection had a tendency to show lower p-values, and nonnegative matrix factorization (NMF) stably showed good performance in many cases unless the Dip test was used for feature selection. In terms of accuracy, the combination of NMF and similarity network fusion (SNF) with Monte Carlo Feature Selection (MCFS) and Minimum-Redundancy Maximum Relevance (mRMR) showed good overall performance. NMF always showed among the worst performances without feature selection in all datasets, but performed much better when used with various feature selection methods. iClusterBayes (ICB) had decent performance when used without feature selection.

CONCLUSIONS

Rather than a single method clearly emerging as optimal, the best methodology was different depending on the data used, the number of features selected, and the evaluation method. A guideline for choosing the best combination method under various situations is provided.

Clustering analysis revealed the autophagy classification and potential autophagy regulators' sensitivity of pancreatic cancer based on multi-omics data

BACKGROUND

Pancreatic ductal adenocarcinoma (PDAC) is a lethal malignancy and is unresponsive to conventional therapeutic modalities due to its high heterogeneity, expounding the necessity, and priority of searching for effective biomarkers and drugs. Autophagy, as an evolutionarily conserved biological process, is upregulated in PDAC and its regulation is linked to a poor prognosis. Increased autophagy sequestered MHC-I on PDAC cells and weaken the antigen presentation and antitumor immune response, indicating the potential therapeutic strategies of autophagy inhibitors.

METHODS

By performing 10 state-of-the-art multi-omics clustering algorithms, we constructed a robust PDAC classification model to reveal the autophagy-related genes among different subgroups.

OUTCOMES

After building a more comprehensive regulating network for potential autophagy regulators exploration, we concluded the top 20 autophagy-related hub genes (GAPDH, MAPK3, RHEB, SQSTM1, EIF2S1, RAB5A, CTSD, MAP1LC3B, RAB7A, RAB11A, FADD, CFKN2A, HSP90AB1, VEGFA, RELA, DDIT3, HSPA5, BCL2L1, BAG3, and ERBB2), six miRNAs, five transcription factors, and five immune infiltrated cells as biomarkers. The drug sensitivity database was screened based on the biomarkers to predict possible drug-targeting signal pathways, hoping to yield novel insights, and promote the progress of the anticancer therapeutic strategy.

CONCLUSION

We succefully constructed an autophagy-related mRNA/miRNA/TF/Immune cells network based on a 10 state-of art algorithm multi-omics analysis, and screened the drug sensitivity dataset for detecting potential signal pathway which might be possible autophagy modulators' targets.

Randomized singular value decomposition for integrative subtype analysis of 'omics data' using non-negative matrix factorization

Integration of multiple 'omics datasets for differentiating cancer subtypes is a powerful technic that leverages the consistent and complementary information across multi-omics data. Matrix factorization is a common technique used in integrative clustering for identifying latent subtype structure across multi-omics data. High dimensionality of the omics data and long computation time have been common challenges of clustering methods. In order to address the challenges, we propose randomized singular value decomposition (RSVD) for integrative clustering using Non-negative Matrix Factorization: intNMF-rsvd. The method utilizes RSVD to reduce the dimensionality by projecting the data into eigen vector space with user specified lower rank. Then, clustering analysis is carried out by estimating common basis matrix across the projected multi-omics datasets. The performance of the proposed method was assessed using the simulated datasets and compared with six state-of-the-art integrative clustering methods using real-life datasets from The Cancer Genome Atlas Study. intNMF-rsvd was found working efficiently and competitively as compared to standard intNMF and other multi-omics clustering methods. Most importantly, intNMF-rsvd can handle large number of features and significantly reduce the computation time. The identified subtypes can be utilized for further clinical association studies to understand the etiology of the disease.

A guide to multi-omics data collection and integration for translational medicine

The emerging high-throughput technologies have led to the shift in the design of translational medicine projects towards collecting multi-omics patient samples and, consequently, their integrated analysis. However, the complexity of integrating these datasets has triggered new questions regarding the appropriateness of the available computational methods. Currently, there is no clear consensus on the best combination of omics to include and the data integration methodologies required for their analysis. This article aims to guide the design of multi-omics studies in the field of translational medicine regarding the types of omics and the integration method to choose. We review articles that perform the integration of multiple omics measurements from patient samples. We identify five objectives in translational medicine applications: (i) detect disease-associated molecular patterns, (ii) subtype identification, (iii) diagnosis/prognosis, (iv) drug response prediction, and (v) understand regulatory processes. We describe common trends in the selection of omic types combined for different objectives and diseases. To guide the choice of data integration tools, we group them into the scientific objectives they aim to address. We describe the main computational methods adopted to achieve these objectives and present examples of tools. We compare tools based on how they deal with the computational challenges of data integration and comment on how they perform against predefined objective-specific evaluation criteria. Finally, we discuss examples of tools for downstream analysis and further extraction of novel insights from multi-omics datasets.

Multi-omics consensus portfolio to refine the classification of lung adenocarcinoma with prognostic stratification, tumor microenvironment, and unique sensitivity to first-line therapies

BACKGROUND

Molecular classification of lung adenocarcinoma (LUAD) based on transcriptomic features has been widely studied. The complementarity of data obtained from multilayer molecular biology could help the LUAD classification via combining multi-omics information.

METHODS

We successfully divided samples from the The Cancer Genome Atlas (TCGA) (n=437) into four subtypes (CS1, CS2, CS3 and CS4) by 10 comprehensive multi-omics clustering methods in the "movics" R package. Meanwhile, external validation sets from different sequencing technologies proved the robustness of the grouping model. The relationship between subtypes, prognosis, molecular features, tumor microenvironment and response to first-line therapy was further analyzed. Next we used univariate Cox regression analysis and Lasso regression analysis to explore the application of biomarkers in clinical prognosis and constructed a prognostic model.

RESULTS

CS1 showed the worst overall survival (OS) among all four clusters, possibly related to its poor immune infiltration, higher tumor mutation and worse chromosomal stability. Patients in different subtypes differed significantly in cancer stem cell characteristics, activation of cancer-related pathways, sensitivity to chemotherapy and immunotherapy. The prognostic model showed good predictive performance. The 1-, 2- and 3-year areas under the curve of risk score were 0.779, 0.742 and 0.678, respectively. Seven genes (DKK1, TSPAN7, ID1, DLGAP5, HHIPL2, CD40 and SEMA3C) used to build the model may be potential therapeutic targets for LUAD.

CONCLUSIONS

Four LUAD subtypes with different molecular characteristics and clinical implications were identified successfully through bioinformatic analysis. Our results may contribute to precision medicine and inform the development of rational clinical strategies for targeted and immune therapies.

A multi-omic analysis of MCF10A cells provides a resource for integrative assessment of ligand-mediated molecular and phenotypic responses

The phenotype of a cell and its underlying molecular state is strongly influenced by extracellular signals, including growth factors, hormones, and extracellular matrix proteins. While these signals are normally tightly controlled, their dysregulation leads to phenotypic and molecular states associated with diverse diseases. To develop a detailed understanding of the linkage between molecular and phenotypic changes, we generated a comprehensive dataset that catalogs the transcriptional, proteomic, epigenomic and phenotypic responses of MCF10A mammary epithelial cells after exposure to the ligands EGF, HGF, OSM, IFNG, TGFB and BMP2. Systematic assessment of the molecular and cellular phenotypes induced by these ligands comprise the LINCS Microenvironment (ME) perturbation dataset, which has been curated and made publicly available for community-wide analysis and development of novel computational methods ( synapse.org/LINCS_MCF10A ). In illustrative analyses, we demonstrate how this dataset can be used to discover functionally related molecular features linked to specific cellular phenotypes. Beyond these analyses, this dataset will serve as a resource for the broader scientific community to mine for biological insights, to compare signals carried across distinct molecular modalities, and to develop new computational methods for integrative data analysis.

Multi-omics subtyping of hepatocellular carcinoma patients using a Bayesian network mixture model

Comprehensive molecular characterization of cancer subtypes is essential for predicting clinical outcomes and searching for personalized treatments. We present bnClustOmics, a statistical model and computational tool for multi-omics unsupervised clustering, which serves a dual purpose: Clustering patient samples based on a Bayesian network mixture model and learning the networks of omics variables representing these clusters. The discovered networks encode interactions among all omics variables and provide a molecular characterization of each patient subgroup. We conducted simulation studies that demonstrated the advantages of our approach compared to other clustering methods in the case where the generative model is a mixture of Bayesian networks. We applied bnClustOmics to a hepatocellular carcinoma (HCC) dataset comprising genome (mutation and copy number), transcriptome, proteome, and phosphoproteome data. We identified three main HCC subtypes together with molecular characteristics, some of which are associated with survival even when adjusting for the clinical stage. Cluster-specific networks shed light on the links between genotypes and molecular phenotypes of samples within their respective clusters and suggest targets for personalized treatments.

A benchmark study of deep learning-based multi-omics data fusion methods for cancer

BACKGROUND

A fused method using a combination of multi-omics data enables a comprehensive study of complex biological processes and highlights the interrelationship of relevant biomolecules and their functions. Driven by high-throughput sequencing technologies, several promising deep learning methods have been proposed for fusing multi-omics data generated from a large number of samples.

RESULTS

In this study, 16 representative deep learning methods are comprehensively evaluated on simulated, single-cell, and cancer multi-omics datasets. For each of the datasets, two tasks are designed: classification and clustering. The classification performance is evaluated by using three benchmarking metrics including accuracy, F1 macro, and F1 weighted. Meanwhile, the clustering performance is evaluated by using four benchmarking metrics including the Jaccard index (JI), C-index, silhouette score, and Davies Bouldin score. For the cancer multi-omics datasets, the methods' strength in capturing the association of multi-omics dimensionality reduction results with survival and clinical annotations is further evaluated. The benchmarking results indicate that moGAT achieves the best classification performance. Meanwhile, efmmdVAE, efVAE, and lfmmdVAE show the most promising performance across all complementary contexts in clustering tasks.

CONCLUSIONS

Our benchmarking results not only provide a reference for biomedical researchers to choose appropriate deep learning-based multi-omics data fusion methods, but also suggest the future directions for the development of more effective multi-omics data fusion methods. The deep learning frameworks are available at https://github.com/zhenglinyi/DL-mo .

Exploration of the Immunotyping Landscape and Immune Infiltration-Related Prognostic Markers in Ovarian Cancer Patients

Background

Increasing evidence indicates that immune cell infiltration (ICI) affects the prognosis of multiple cancers. This study aims to explore the immunotypes and ICI-related biomarkers in ovarian cancer.

Methods

The ICI levels were quantified with the CIBERSORT and ESTIMATE algorithms. The unsupervised consensus clustering method determined immunotypes based on the ICI profiles. Characteristic genes were identified with the Boruta algorithm. Then, the ICI score, a novel prognostic marker, was generated with the principal component analysis of the characteristic genes. The relationships between the ICI scores and clinical features were revealed. Further, an ICI signature was integrated after the univariate Cox, lasso, and stepwise regression analyses. The accuracy and robustness of the model were tested by three independent cohorts. The roles of the model in the immunophenoscores (IPS), tumor immune dysfunction and exclusion (TIDE) scores, and immunotherapy responses were also explored. Finally, risk genes (GBP1P1, TGFBI, PLA2G2D) and immune cell marker genes (CD11B, NOS2, CD206, CD8A) were tested by qRT-PCR in clinical tissues.

Results

Three immunotypes were identified, and ICI scores were generated based on the 75 characteristic genes. CD8 TCR pathways, chemokine-related pathways, and lymphocyte activation were critical to immunophenotyping. Higher ICI scores contributed to better prognoses. An independent prognostic factor, a three-gene signature, was integrated to calculate patients’ risk scores. Higher TIDE scores, lower ICI scores, lower IPS, lower immunotherapy responses, and worse prognoses were revealed in high-risk patients. Macrophage polarization and CD8 T cell infiltration were indicated to play potentially important roles in the development of ovarian cancer in the clinical validation cohort.

Conclusions

Our study characterized the immunotyping landscape and provided novel immune infiltration-related prognostic markers in ovarian cancer.

Integrative clustering methods for multi-omics data

Integrative analysis of multi-omics data has drawn much attention from the scientific community due to the technological advancements which have generated various omics data. Leveraging these multi-omics data potentially provides a more comprehensive view of the disease mechanism or biological processes. Integrative multi-omics clustering is an unsupervised integrative method specifically used to find coherent groups of samples or features by utilizing information across multi-omics data. It aims to better stratify diseases and to suggest biological mechanisms and potential targeted therapies for the diseases. However, applying integrative multi-omics clustering is both statistically and computationally challenging due to various reasons such as high dimensionality and heterogeneity. In this review, we summarized integrative multi-omics clustering methods into three general categories: concatenated clustering, clustering of clusters, and interactive clustering based on when and how the multi-omics data are processed for clustering. We further classified the methods into different approaches under each category based on the main statistical strategy used during clustering. In addition, we have provided recommended practices tailored to four real-life scenarios to help researchers to strategize their selection in integrative multi-omics clustering methods for their future studies.

Unsupervised Multi-Omics Data Integration Methods: A Comprehensive Review

Through the developments of Omics technologies and dissemination of large-scale datasets, such as those from The Cancer Genome Atlas, Alzheimer’s Disease Neuroimaging Initiative, and Genotype-Tissue Expression, it is becoming increasingly possible to study complex biological processes and disease mechanisms more holistically. However, to obtain a comprehensive view of these complex systems, it is crucial to integrate data across various Omics modalities, and also leverage external knowledge available in biological databases. This review aims to provide an overview of multi-Omics data integration methods with different statistical approaches, focusing on unsupervised learning tasks, including disease onset prediction, biomarker discovery, disease subtyping, module discovery, and network/pathway analysis. We also briefly review feature selection methods, multi-Omics data sets, and resources/tools that constitute critical components for carrying out the integration.

PIntMF: Penalized Integrative Matrix Factorization method for multi-omics data

MOTIVATION

It is more and more common to perform multi-omics analyses to explore the genome at diverse levels and not only at a single level. Through integrative statistical methods, multi-omics data have the power to reveal new biological processes, potential biomarkers and subgroups in a cohort. Matrix factorization (MF) is an unsupervised statistical method that allows a clustering of individuals, but also reveals relevant omics variables from the various blocks.

RESULTS

Here, we present PIntMF (Penalized Integrative Matrix Factorization), an MF model with sparsity, positivity and equality constraints. To induce sparsity in the model, we used a classical Lasso penalization on variable and individual matrices. For the matrix of samples, sparsity helps in the clustering, while normalization (matching an equality constraint) of inferred coefficients is added to improve interpretation. Moreover, we added an automatic tuning of the sparsity parameters using the famous glmnet package. We also proposed three criteria to help the user to choose the number of latent variables. PIntMF was compared with other state-of-the-art integrative methods including feature selection techniques in both synthetic and real data. PIntMF succeeds in finding relevant clusters as well as variables in two types of simulated data (correlated and uncorrelated). Next, PIntMF was applied to two real datasets (Diet and cancer), and it revealed interpretable clusters linked to available clinical data. Our method outperforms the existing ones on two criteria (clustering and variable selection). We show that PIntMF is an easy, fast and powerful tool to extract patterns and cluster samples from multi-omics data.

AVAILABILITY AND IMPLEMENTATION

An R package is available at https://github.com/mpierrejean/pintmf.

SUPPLEMENTARY INFORMATION

Supplementary data are available at Bioinformatics online.

Genome-scale metabolic modeling reveals SARS-CoV-2-induced metabolic changes and antiviral targets

Tremendous progress has been made to control the COVID-19 pandemic caused by the SARS-CoV-2 virus. However, effective therapeutic options are still rare. Drug repurposing and combination represent practical strategies to address this urgent unmet medical need. Viruses, including coronaviruses, are known to hijack host metabolism to facilitate viral proliferation, making targeting host metabolism a promising antiviral approach. Here, we describe an integrated analysis of 12 published in vitro and human patient gene expression datasets on SARS-CoV-2 infection using genome-scale metabolic modeling (GEM), revealing complicated host metabolism reprogramming during SARS-CoV-2 infection. We next applied the GEM-based metabolic transformation algorithm to predict anti-SARS-CoV-2 targets that counteract the virus-induced metabolic changes. We successfully validated these targets using published drug and genetic screen data and by performing an siRNA assay in Caco-2 cells. Further generating and analyzing RNA-sequencing data of remdesivir-treated Vero E6 cell samples, we predicted metabolic targets acting in combination with remdesivir, an approved anti-SARS-CoV-2 drug. Our study provides clinical data-supported candidate anti-SARS-CoV-2 targets for future evaluation, demonstrating host metabolism targeting as a promising antiviral strategy.

Genome-scale metabolic modeling reveals SARS-CoV-2-induced metabolic changes and antiviral targets

Tremendous progress has been made to control the COVID-19 pandemic caused by the SARS-CoV-2 virus. However, effective therapeutic options are still rare. Drug repurposing and combination represent practical strategies to address this urgent unmet medical need. Viruses, including coronaviruses, are known to hijack host metabolism to facilitate viral proliferation, making targeting host metabolism a promising antiviral approach. Here, we describe an integrated analysis of 12 published in vitro and human patient gene expression datasets on SARS-CoV-2 infection using genome-scale metabolic modeling (GEM), revealing complicated host metabolism reprogramming during SARS-CoV-2 infection. We next applied the GEM-based metabolic transformation algorithm to predict anti-SARS-CoV-2 targets that counteract the virus-induced metabolic changes. We successfully validated these targets using published drug and genetic screen data and by performing an siRNA assay in Caco-2 cells. Further generating and analyzing RNA-sequencing data of remdesivir-treated Vero E6 cell samples, we predicted metabolic targets acting in combination with remdesivir, an approved anti-SARS-CoV-2 drug. Our study provides clinical data-supported candidate anti-SARS-CoV-2 targets for future evaluation, demonstrating host metabolism-targeting as a promising antiviral strategy.

A Hierarchical Graph Convolution Network for Representation Learning of Gene Expression Data

The curse of dimensionality, which is caused by high-dimensionality and low-sample-size, is a major challenge in gene expression data analysis. However, the real situation is even worse: labelling data is laborious and time-consuming, so only a small part of the limited samples will be labelled. Having such few labelled samples further increases the difficulty of training deep learning models. Interpretability is an important requirement in biomedicine. Many existing deep learning methods are trying to provide interpretability, but rarely apply to gene expression data. Recent semi-supervised graph convolution network methods try to address these problems by smoothing the label information over a graph. However, to the best of our knowledge, these methods only utilize graphs in either the feature space or sample space, which restrict their performance. We propose a transductive semi-supervised representation learning method called a hierarchical graph convolution network (HiGCN) to aggregate the information of gene expression data in both feature and sample spaces. HiGCN first utilizes external knowledge to construct a feature graph and a similarity kernel to construct a sample graph. Then, two spatial-based GCNs are used to aggregate information on these graphs. To validate the model's performance, synthetic and real datasets are provided to lend empirical support. Compared with two recent models and three traditional models, HiGCN learns better representations of gene expression data, and these representations improve the performance of downstream tasks, especially when the model is trained on a few labelled samples. Important features can be extracted from our model to provide reliable interpretability.

Integration strategies of multi-omics data for machine learning analysis

Increased availability of high-throughput technologies has generated an ever-growing number of omics data that seek to portray many different but complementary biological layers including genomics, epigenomics, transcriptomics, proteomics, and metabolomics. New insight from these data have been obtained by machine learning algorithms that have produced diagnostic and classification biomarkers. Most biomarkers obtained to date however only include one omic measurement at a time and thus do not take full advantage of recent multi-omics experiments that now capture the entire complexity of biological systems.

Multi-omics data integration strategies are needed to combine the complementary knowledge brought by each omics layer. We have summarized the most recent data integration methods/ frameworks into five different integration strategies: early, mixed, intermediate, late and hierarchical. In this mini-review, we focus on challenges and existing multi-omics integration strategies by paying special attention to machine learning applications.

Undisclosed, unmet and neglected challenges in multi-omics studies

Multi-omics approaches have become a reality in both large genomics projects and small laboratories. However, the multi-omics research community still faces a number of issues that have either not been sufficiently discussed or for which current solutions are still limited. In this Perspective, we elaborate on these limitations and suggest points of attention for future research. We finally discuss new opportunities and challenges brought to the field by the rapid development of single-cell high-throughput molecular technologies.

TSCCA: A tensor sparse CCA method for detecting microRNA-gene patterns from multiple cancers

Existing studies have demonstrated that dysregulation of microRNAs (miRNAs or miRs) is involved in the initiation and progression of cancer. Many efforts have been devoted to identify microRNAs as potential biomarkers for cancer diagnosis, prognosis and therapeutic targets. With the rapid development of miRNA sequencing technology, a vast amount of miRNA expression data for multiple cancers has been collected. These invaluable data repositories provide new paradigms to explore the relationship between miRNAs and cancer. Thus, there is an urgent need to explore the complex cancer-related miRNA-gene patterns by integrating multi-omics data in a pan-cancer paradigm. In this study, we present a tensor sparse canonical correlation analysis (TSCCA) method for identifying cancer-related miRNA-gene modules across multiple cancers. TSCCA is able to overcome the drawbacks of existing solutions and capture both the cancer-shared and specific miRNA-gene co-expressed modules with better biological interpretations. We comprehensively evaluate the performance of TSCCA using a set of simulated data and matched miRNA/gene expression data across 33 cancer types from the TCGA database. We uncover several dysfunctional miRNA-gene modules with important biological functions and statistical significance. These modules can advance our understanding of miRNA regulatory mechanisms of cancer and provide insights into miRNA-based treatments for cancer.

MicroRNAs (miRNAs) are a class of small non-coding RNAs. Previous studies have revealed that miRNA-gene regulatory modules play key roles in the occurrence and development of cancer. However, little has been done to discover miRNA-gene regulatory modules from a pan-cancer view. Thus, it is urgently needed to develop new methods to explore the complex cancer-related miRNA-gene patterns by integrating multi-omics data of multi-cancers. To build the connections between miRNA-gene regulatory modules across different cancer types, we propose a tensor sparse canonical correlation analysis (TSCCA) method. Our specific contributions are two-fold: (1) We propose a sparse statistical learning model TSCCA and an efficient block-coordinate descent algorithm to solve it. (2) We apply TSCCA to a multi-omics data set of 33 cancer types from TCGA and identify some cancer-related miRNA-gene modules with important biological functions and statistical significance.

Molecular Landscape and Therapeutic Strategies in Cholangiocarcinoma: An Integrated Translational Approach towards Precision Medicine

Cholangiocarcinomas (CCAs) are heterogeneous biliary tract malignancies with dismal prognosis, mainly due to tumor aggressiveness, late diagnosis, and poor response to current therapeutic options. High-throughput technologies have been used as a fundamental tool in unveiling CCA molecular landscape, and several molecular classifications have been proposed, leading to various targeted therapy trials. In this review, we aim to analyze the critical issues concerning the status of precision medicine in CCA, discussing molecular signatures and clusters, related to both anatomical classification and different etiopathogenesis, and the latest therapeutic strategies. Furthermore, we propose an integrated approach comprising the CCA molecular mechanism, pathobiology, clinical and histological findings, and treatment perspectives for the ultimate purpose of improving the methods of patient allocations in clinical trials and the response to personalized therapies.

Big data, integrative omics and network biology

A cell integrates various signals through a network of biomolecules that crosstalk to synergistically regulate the replication, transcription, translation and other metabolic activities of a cell. These networks regulate signal perception and processing that drives biological functions. The biological complexity cannot be fully captured by a single -omics discipline. The holistic study of an organism-in health, perturbation, exposure to environment and disease, is studied under systems biology. The bottom-up molecular approaches (genes, mRNA, protein, metabolite, etc.) have laid the foundation of current biological knowledge covering the horizon from viruses, bacteria, fungi, plants and animals. Yet, these techniques provide a rather myopic view of biology at the molecular level. To understand how the interconnected molecular components are formed and rewired in disease or exposure to environmental stimuli is the holy grail of modern biology. The omics era was heralded by the genomics revolution but advanced sequencing techniques are now also ubiquitous in transcriptomics, proteomics, metabolomics and lipidomics. Multi-omics data analysis and integration techniques are driving the quest for deeper insights into how the different layers of biomolecules talk to each other in diverse contexts.

DNA methylation and gene expression integration in cardiovascular disease

BACKGROUND

The integration of different layers of omics information is an opportunity to tackle the complexity of cardiovascular diseases (CVD) and to identify new predictive biomarkers and potential therapeutic targets. Our aim was to integrate DNA methylation and gene expression data in an effort to identify biomarkers related to cardiovascular disease risk in a community-based population. We accessed data from the Framingham Offspring Study, a cohort study with data on DNA methylation (Infinium HumanMethylation450 BeadChip; Illumina) and gene expression (Human Exon 1.0 ST Array; Affymetrix). Using the MOFA2 R package, we integrated these data to identify biomarkers related to the risk of presenting a cardiovascular event.

RESULTS

Four independent latent factors (9, 19, 21-only in women-and 27), driven by DNA methylation, were associated with cardiovascular disease independently of classical risk factors and cell-type counts. In a sensitivity analysis, we also identified factor 21 as associated with CVD in women. Factors 9, 21 and 27 were also associated with coronary heart disease risk. Moreover, in a replication effort in an independent study three of the genes included in factor 27 were also present in a factor identified to be associated with myocardial infarction (CDC42BPB, MAN2A2 and RPTOR). Factor 9 was related to age and cell-type proportions; factor 19 was related to age and B cells count; factor 21 pointed to human immunodeficiency virus infection-related pathways and inflammation; and factor 27 was related to lifestyle factors such as alcohol consumption, smoking and body mass index. Inclusion of factor 21 (only in women) improved the discriminative and reclassification capacity of the Framingham classical risk function and factor 27 improved its discrimination.

CONCLUSIONS

Unsupervised multi-omics data integration methods have the potential to provide insights into the pathogenesis of cardiovascular diseases. We identified four independent factors (one only in women) pointing to inflammation, endothelium homeostasis, visceral fat, cardiac remodeling and lifestyles as key players in the determination of cardiovascular risk. Moreover, two of these factors improved the predictive capacity of a classical risk function.

Eight-lncRNA signature of cervical cancer were identified by integrating DNA methylation, copy number variation and transcriptome data

BACKGROUND

Copy number variation (CNV) suggests genetic changes in malignant tumors. Abnormal expressions of long non-coding RNAs (lncRNAs) resulted from genomic and epigenetic abnormalities play a driving role in tumorigenesis of cervical cancer. However, the role of lncRNAs-related CNV in cervical cancer remained largely unclear.

METHODS

The data of messenger RNAs (mRNAs), DNA methylation, and DNA copy number were collected from 292 cervical cancer specimens. The prognosis-related subtypes of cervical cancer were determined by multi-omics integration analysis, and protein-coding genes (PCGs) and lncRNAs with subtype-specific expressions were identified. The CNV pattern of the subtype-specific lncRNAs was analyzed to identify the subtype-specific lncRNAs. A prognostic risk model based on lncRNAs was established by least absolute shrinkage and selection operator (LASSO).

RESULTS

Multi-omics integration analysis identified three molecular subtypes incorporating 617 differentially expressed lncRNAs and 1395 differentially expressed PCGs. The 617 lncRNAs were found to intersect with disease-related lncRNAs. Functional enrichment showed that 617 lncRNAs were mainly involved in tumor metabolism, immunity and other pathways, such as p53 and cAMP signaling pathways, which are closely related to the development of cervical cancer. Finally, according to CNV pattern consistent with differential expression analysis, we established a lncRNAs-based signature consisted of 8 lncRNAs, namely, RUSC1-AS1, LINC01990, LINC01411, LINC02099, H19, LINC00452, ADPGK-AS1, C1QTNF1-AS1. The interaction of the 8 lncRNAs showed a significantly poor prognosis of cervical cancer patients, which has also been verified in an independent dataset.

CONCLUSION

Our study expanded the network of CNVs and improved the understanding on the regulatory network of lncRNAs in cervical cancer, providing novel biomarkers for the prognosis management of cervical cancer patients.

Benchmarking joint multi-omics dimensionality reduction approaches for the study of cancer

High-dimensional multi-omics data are now standard in biology. They can greatly enhance our understanding of biological systems when effectively integrated. To achieve proper integration, joint Dimensionality Reduction (jDR) methods are among the most efficient approaches. However, several jDR methods are available, urging the need for a comprehensive benchmark with practical guidelines. We perform a systematic evaluation of nine representative jDR methods using three complementary benchmarks. First, we evaluate their performances in retrieving ground-truth sample clustering from simulated multi-omics datasets. Second, we use TCGA cancer data to assess their strengths in predicting survival, clinical annotations and known pathways/biological processes. Finally, we assess their classification of multi-omics single-cell data. From these in-depth comparisons, we observe that intNMF performs best in clustering, while MCIA offers an effective behavior across many contexts. The code developed for this benchmark study is implemented in a Jupyter notebook—multi-omics mix (momix)—to foster reproducibility, and support users and future developers.

Advances in omics technology have resulted in the generation of multi-view data for cancer samples. Here, the authors compare dimensionality reduction techniques using simulated and TCGA data and identify the features of the methods with superior performance.

A Review and a Framework of Variables for Defining and Characterizing Tinnitus Subphenotypes

Tinnitus patients can present with various characteristics, such as those related to the tinnitus perception, symptom severity, and pattern of comorbidities. It is speculated that this phenotypic heterogeneity is associated with differences in the underlying pathophysiology and personal reaction to the condition. However, there is as yet no established protocol for tinnitus profiling or subtyping, hindering progress in treatment development. This review summarizes data on variables that have been used in studies investigating phenotypic differences in subgroups of tinnitus, including variables used to both define and compare subgroups. A PubMed search led to the identification of 64 eligible articles. In most studies, variables for subgrouping were chosen by the researchers (hypothesis-driven approach). Other approaches included application of unsupervised machine-learning techniques for the definition of subgroups (data-driven), and subgroup definition based on the response to a tinnitus treatment (treatment response). A framework of 94 variable concepts was created to summarize variables used across all studies. Frequency statistics for the use of each variable concept are presented, demonstrating those most and least commonly assessed. This review highlights the high dimensionality of tinnitus heterogeneity. The framework of variables can contribute to the design of future studies, helping to decide on tinnitus assessment and subgrouping.