Towards multi-omics characterization of tumor heterogeneity: a comprehensive review of statistical and machine learning approaches

Integrated multi-omics analysis and machine learning to refine molecular subtypes, prognosis, and immunotherapy in lung adenocarcinoma

Lung adenocarcinoma (LUAD) has a malignant characteristic that is highly aggressive and prone to metastasis. There is still a lack of suitable biomarkers to facilitate the refinement of precision-based therapeutic regimens. We used a combination of 10 known clustering algorithms and the omics data from 4 dimensions to identify high-resolution molecular subtypes of LUAD. Subsequently, consensus machine learning-related prognostic signature (CMRS) was developed based on subtypes related genes and an integrated program framework containing 10 machine learning algorithms. The efficiency of CMRS was analyzed from the perspectives of tumor microenvironment, genomic landscape, immunotherapy, drug sensitivity, and single-cell analysis. In terms of results, through multi-omics clustering, we identified 2 comprehensive omics subtypes (CSs) in which CS1 patients had worse survival outcomes, higher aggressiveness, mRNAsi and mutation frequency. Subsequently, we developed CMRS based on 13 key genes up-regulated in CS1. The prognostic predictive efficiency of CMRS was superior to most established LUAD prognostic signatures. CMRS demonstrated a strong correlation with tumor microenvironmental feature variants and genomic instability generation. Regarding clinical performance, patients in the high CMRS group were more likely to benefit from immunotherapy, whereas low CMRS were more likely to benefit from chemotherapy and targeted drug therapy. In addition, we evaluated that drugs such as neratinib, oligomycin A, and others may be candidates for patients in the high CMRS group. Single-cell analysis revealed that CMRS-related genes were mainly expressed in epithelial cells. The novel molecular subtypes identified in this study based on multi-omics data could provide new insights into the stratified treatment of LUAD, while the development of CMRS could serve as a candidate indicator of the degree of benefit of precision therapy and immunotherapy for LUAD.

Omics-based molecular classifications empowering in precision oncology

BACKGROUND

In the past decades, cancer enigmatical heterogeneity at distinct expression levels could interpret disparities in therapeutic response and prognosis. It built hindrances to precision medicine, a tactic to tailor customized treatment informed by the tumors' molecular profile. Single-omics analysis dissected the biological features associated with carcinogenesis to some extent but still failed to revolutionize cancer treatment as expected. Integrated omics analysis incorporated tumor biological networks from diverse layers and deciphered a holistic overview of cancer behaviors, yielding precise molecular classification to facilitate the evolution and refinement of precision medicine.

CONCLUSION

This review outlined the biomarkers at multiple expression layers to tutor molecular classification and pinpoint tumor diagnosis, and explored the paradigm shift in precision therapy: from single- to multi-omics-based subtyping to optimize therapeutic regimens. Ultimately, we firmly believe that by parsing molecular characteristics, omics-based typing will be a powerful assistant for precision oncology.

Computational frameworks integrating deep learning and statistical models in mining multimodal omics data

BACKGROUND

In health research, multimodal omics data analysis is widely used to address important clinical and biological questions. Traditional statistical methods rely on the strong assumptions of distribution. Statistical methods such as testing and differential expression are commonly used in omics analysis. Deep learning, on the other hand, is an advanced computer science technique that is powerful in mining high-dimensional omics data for prediction tasks. Recently, integrative frameworks or methods have been developed for omics studies that combine statistical models and deep learning algorithms.

METHODS AND RESULTS

The aim of these integrative frameworks is to combine the strengths of both statistical methods and deep learning algorithms to improve prediction accuracy while also providing interpretability and explainability. This review report discusses the current state-of-the-art integrative frameworks, their limitations, and potential future directions in survival and time-to-event longitudinal analysis, dimension reduction and clustering, regression and classification, feature selection, and causal and transfer learning.

Factors affecting heterogeneity in breast cancer microenvironment: A narrative mini review

Breast cancer (BC) heterogeneity is a key trait of BC tumors with crucial implications on tumorigenesis, diagnosis, and therapeutic modalities. It is influenced by tumor intrinsic features and by the tumor microenvironment (TME) composition of different intra-tumoral regions, which in turn affect cancer progression within patients. In this mini review, we will highlight the mechanisms that generate cancer heterogeneity in BC and how they affect the responses to cancer therapies.

Standardizing analysis of intra-tumoral heterogeneity with computational pathology

Many malignant cancers like glioblastoma are highly adaptive diseases that dynamically change their regional biology to survive and thrive under diverse microenvironmental and therapeutic pressures. While the concept of intra-tumoral heterogeneity has become a major paradigm in cancer research and care, systematic approaches to assess and document bio-variation in cancer are still in their infancy. Here we discuss existing approaches and challenges to documenting intra-tumoral heterogeneity and emerging computational approaches that leverage artificial intelligence to begin to overcome these limitations. We propose how these emerging techniques can be coupled with a diversity of molecular tools to address intra-tumoral heterogeneity more systematically in research and in practice, especially across larger specimens and longitudinal analyses. Systematic documentation and characterization of heterogeneity across entire tumor specimens and their longitudinal evolution has the potential to improve our understanding and treatment of cancer.

Unravelling the Complexity of Colorectal Cancer: Heterogeneity, Clonal Evolution, and Clinical Implications

Colorectal cancer (CRC) is a global health concern and a leading cause of death worldwide. The disease's course and response to treatment are significantly influenced by its heterogeneity, both within a single lesion and between primary and metastatic sites. Biomarkers, such as mutations in KRAS, NRAS, and BRAF, provide valuable guidance for treatment decisions in patients with metastatic CRC. While high concordance exists between mutational status in primary and metastatic lesions, some heterogeneity may be present. Circulating tumor DNA (ctDNA) analysis has proven invaluable in identifying genetic heterogeneity and predicting prognosis in RAS-mutated metastatic CRC patients. Tumor heterogeneity can arise from genetic and non-genetic factors, affecting tumor development and response to therapy. To comprehend and address clonal evolution and intratumoral heterogeneity, comprehensive genomic studies employing techniques such as next-generation sequencing and computational analysis are essential. Liquid biopsy, notably through analysis of ctDNA, enables real-time clonal evolution and treatment response monitoring. However, challenges remain in standardizing procedures and accurately characterizing tumor subpopulations. Various models elucidate the origin of CRC heterogeneity, highlighting the intricate molecular pathways involved. This review focuses on intrapatient cancer heterogeneity and genetic clonal evolution in metastatic CRC, with an emphasis on clinical applications.

Machine Learning in Metastatic Cancer Research: Potentials, Possibilities, and Prospects

Cancer has received extensive recognition for its high mortality rate, with metastatic cancer being the top cause of cancer-related deaths. Metastatic cancer involves the spread of the primary tumor to other body organs. As much as the early detection of cancer is essential, the timely detection of metastasis, the identification of biomarkers, and treatment choice are valuable for improving the quality of life for metastatic cancer patients. This study reviews the existing studies on classical machine learning (ML) and deep learning (DL) in metastatic cancer research. Since the majority of metastatic cancer research data are collected in the formats of PET/CT and MRI image data, deep learning techniques are heavily involved. However, its black-box nature and expensive computational cost are notable concerns. Furthermore, existing models could be overestimated for their generality due to the non-diverse population in clinical trial datasets. Therefore, research gaps are itemized; follow-up studies should be carried out on metastatic cancer using machine learning and deep learning tools with data in a symmetric manner.

Genome-wide transcriptomics and copy number profiling identify patient-specific CNV-lncRNA-mRNA regulatory triplets in colorectal cancer

Screening cancer genomes has provided an in-depth characterization of genetic variants such as copy number variations (CNVs) and gene expression changes of non-coding transcripts. Single-dimensional experiments are often designed to differentiate a patient cohort into various sets with the aim of identifying molecular changes among groups; however, this may be inadequate to decipher the causal relationship between molecular signatures in individual patients. To overcome this challenge with respect to personalized medicine, we implemented a patient-specific multi-dimensional integrative approach to uncover coherent signals from multiple independent platforms. In particular, we focused on the consistent gene dosage effects of CNVs for both mRNA and long non-coding RNA (lncRNA) expression in nine colorectal cancer patients. We identified 511 CNV-lncRNA-mRNA regulatory triplets associated with CNVs and aberrant expression of both mRNAs and lncRNAs. By filtering out inconsistent changes among CNVs, mRNAs, and lncRNAs, we further characterized 165 coherent motifs associated with 56 genes. In total, 108 motifs were linked with 31 copy number gains, 44 upregulated lncRNAs, and 45 upregulated mRNAs. Another 57 coherent downregulated motifs were also collected. We discuss how for many of these CNV-lncRNA-mRNA regulatory triplets, their clinical impact remains to be explored, including survival time, microsatellite instability, tumor stage, and primary tumor sites. By validating two example CNV-lncRNA-mRNA triplets with up- and down-regulation, we confirmed that individual variations in multiple dimensions are a robust tool to identify reliable molecular signals for personalized medicine. In summary, we utilized a patient-specific computational pipeline to explore the consistent CNV-driven motifs consisting of lncRNAs and mRNAs. We also identified LSM14B as a potential promoter in colorectal cancer progression, suggesting that it may serve as a target for colorectal cancer treatment.

A Boolean-based machine learning framework identifies predictive biomarkers of HSP90-targeted therapy response in prostate cancer

Precision medicine has emerged as an important paradigm in oncology, driven by the significant heterogeneity of individual patients' tumour. A key prerequisite for effective implementation of precision oncology is the development of companion biomarkers that can predict response to anti-cancer therapies and guide patient selection for clinical trials and/or treatment. However, reliable predictive biomarkers are currently lacking for many anti-cancer therapies, hampering their clinical application. Here, we developed a novel machine learning-based framework to derive predictive multi-gene biomarker panels and associated expression signatures that accurately predict cancer drug sensitivity. We demonstrated the power of the approach by applying it to identify response biomarker panels for an Hsp90-based therapy in prostate cancer, using proteomic data profiled from prostate cancer patient-derived explants. Our approach employs a rational feature section strategy to maximise model performance, and innovatively utilizes Boolean algebra methods to derive specific expression signatures of the marker proteins. Given suitable data for model training, the approach is also applicable to other cancer drug agents in different tumour settings.

Artificial intelligence-based multi-omics analysis fuels cancer precision medicine

With biotechnological advancements, innovative omics technologies are constantly emerging that have enabled researchers to access multi-layer information from the genome, epigenome, transcriptome, proteome, metabolome, and more. A wealth of omics technologies, including bulk and single-cell omics approaches, have empowered to characterize different molecular layers at unprecedented scale and resolution, providing a holistic view of tumor behavior. Multi-omics analysis allows systematic interrogation of various molecular information at each biological layer while posing tricky challenges regarding how to extract valuable insights from the exponentially increasing amount of multi-omics data. Therefore, efficient algorithms are needed to reduce the dimensionality of the data while simultaneously dissecting the mysteries behind the complex biological processes of cancer. Artificial intelligence has demonstrated the ability to analyze complementary multi-modal data streams within the oncology realm. The coincident development of multi-omics technologies and artificial intelligence algorithms has fuelled the development of cancer precision medicine. Here, we present state-of-the-art omics technologies and outline a roadmap of multi-omics integration analysis using an artificial intelligence strategy. The advances made using artificial intelligence-based multi-omics approaches are described, especially concerning early cancer screening, diagnosis, response assessment, and prognosis prediction. Finally, we discuss the challenges faced in multi-omics analysis, along with tentative future trends in this field. With the increasing application of artificial intelligence in multi-omics analysis, we anticipate a shifting paradigm in precision medicine becoming driven by artificial intelligence-based multi-omics technologies.

Biomarkers for the Detection and Risk Stratification of Aggressive Prostate Cancer

Current strategies for the clinical management of prostate cancer are inadequate for a precise risk stratification between indolent and aggressive tumors. Recently developed tissue-based molecular biomarkers have refined the risk assessment of the disease. The characterization of tissue biopsy components and subsequent identification of relevant tissue-based molecular alterations have the potential to improve the clinical decision making and patient outcomes. However, tissue biopsies are invasive and spatially restricted due to tumor heterogeneity. Therefore, there is an urgent need for complementary diagnostic and prognostic options. Liquid biopsy approaches are minimally invasive with potential utility for the early detection, risk stratification, and monitoring of tumors. In this review, we focus on tissue and liquid biopsy biomarkers for early diagnosis and risk stratification of prostate cancer, including modifications on the genomic, epigenomic, transcriptomic, and proteomic levels. High-risk molecular alterations combined with orthogonal clinical parameters can improve the identification of aggressive tumors and increase patient survival.

Computational approaches for network-based integrative multi-omics analysis

Advances in omics technologies allow for holistic studies into biological systems. These studies rely on integrative data analysis techniques to obtain a comprehensive view of the dynamics of cellular processes, and molecular mechanisms. Network-based integrative approaches have revolutionized multi-omics analysis by providing the framework to represent interactions between multiple different omics-layers in a graph, which may faithfully reflect the molecular wiring in a cell. Here we review network-based multi-omics/multi-modal integrative analytical approaches. We classify these approaches according to the type of omics data supported, the methods and/or algorithms implemented, their node and/or edge weighting components, and their ability to identify key nodes and subnetworks. We show how these approaches can be used to identify biomarkers, disease subtypes, crosstalk, causality, and molecular drivers of physiological and pathological mechanisms. We provide insight into the most appropriate methods and tools for research questions as showcased around the aetiology and treatment of COVID-19 that can be informed by multi-omics data integration. We conclude with an overview of challenges associated with multi-omics network-based analysis, such as reproducibility, heterogeneity, (biological) interpretability of the results, and we highlight some future directions for network-based integration.

ADH1C inhibits progression of colorectal cancer through the ADH1C/PHGDH /PSAT1/serine metabolic pathway

Colorectal cancer (CRC) is the third most common cancer in men and the second most common cancer in women worldwide. CRC is the second leading cause of cancer-related deaths. Although some progress in the treatment of CRC has been achieved, the molecular mechanism of CRC is still unclear. In this study, alcohol dehydrogenase 1C(ADH1C) was first identified as a target gene closely associated with the development of CRC by the comprehensive application of transcriptomics, proteomics, metabonomics and in silico analysis. The ADH1C mRNA and protein expression in CRC cell lines and tumor tissues was lower than that in normal intestinal epithelial cell lines and healthy tissues. Overexpression of ADH1C inhibited the growth, migration, invasion and colony formation of CRC cell lines and prevented the growth of xenograft tumors in nude mice. The inhibitory effects of ADH1C on CRC cells in vitro were exerted by reducing the expression of PHGDH/PSAT1 and the serine level. This inhibition could be partially reversed by adding serine to the culture medium. These results showed that ADH1C is a potential drug target in CRC.

Bringing precision oncology to cellular resolution with single-cell genomics

Single-cell sequencing technologies have undergone rapid development and adoption by the scientific community in the past 5 years, fueling discoveries about the etiology, pathogenesis, and treatment responsiveness of individual tumor cells within cancer ecosystems. Most of the advancements in our understanding of cancer with these new technologies have focused on basic tumor biology. However, the knowledge produced by these and other studies are beginning to provide biomarkers and drug targets for clinically-relevant subpopulations within a tumor, creating opportunities for the development of biologically-informed, clone-specific combination treatment strategies. Here we provide an overview of the development of the field of single-cell cancer sequencing and provide a roadmap for shepherding these technologies from research tools to diagnostic instruments that provide high-resolution, treatment-directing details of tumors to clinical oncologists.

Omics of endothelial cell dysfunction in sepsis

During sepsis, defined as life-threatening organ dysfunction due to dysregulated host response to infection, systemic inflammation activates endothelial cells and initiates a multifaceted cascade of pro-inflammatory signaling events, resulting in increased permeability and excessive recruitment of leukocytes. Vascular endothelial cells share many common properties but have organ-specific phenotypes with unique structure and function. Thus, therapies directed against endothelial cell phenotypes are needed to address organ-specific endothelial cell dysfunction. Omics allow for the study of expressed genes, proteins and/or metabolites in biological systems and provide insight on temporal and spatial evolution of signals during normal and diseased conditions. Proteomics quantifies protein expression, identifies protein-protein interactions and can reveal mechanistic changes in endothelial cells that would not be possible to study via reductionist methods alone. In this review, we provide an overview of how sepsis pathophysiology impacts omics with a focus on proteomic analysis of mouse endothelial cells during sepsis/inflammation and its relationship with the more clinically relevant omics of human endothelial cells. We discuss how omics has been used to define septic endotype signatures in different populations with a focus on proteomic analysis in organ-specific microvascular endothelial cells during sepsis or septic-like inflammation. We believe that studies defining septic endotypes based on proteomic expression in endothelial cell phenotypes are urgently needed to complement omic profiling of whole blood and better define sepsis subphenotypes. Lastly, we provide a discussion of how in silico modeling can be used to leverage the large volume of omics data to map response pathways in sepsis.

Multi-Omics analysis identifies a lncRNA-related prognostic signature to predict bladder cancer recurrence

Bladder cancer (BLCA) is one of the most common cancers worldwide with high recurrence rate. Hence, we intended to establish a recurrence-related long non-coding RNA (lncRNA) model of BLCA as a potential biomarker based on multi-omics analysis. Multi-omics data including copy number variation (CNV) data, mutation annotation files, RNA expression profiles and clinical data of The Cancer Genome Atlas (TCGA) BLCA cohort (303 cases) and GSE31684 (93 cases) were downloaded from public database. With multi-omics analysis, twenty lncRNAs were identified as the candidates related with BLCA recurrence, CNVs and mutations in training set. Ten-lncRNA signature were established using least absolute shrinkage and selection operation (LASSO) and Cox regression. Then, various survival analysis was used to assess the power of lncRNA model in predicting BLCA recurrence. The results showed that the recurrence-free survival time of high-risk group was significantly shorter than that of low-risk group in training and testing sets, and the predictive value of ten-lncRNA signature was robust and independent of other clinical variables. Gene Set Enrichment Analysis (GSEA) showed this signature were associated with immune disorders, indicating this signature may be involved in tumor immunology. After compared with the other reported lncRNA signatures, ten-lncRNA signature was validated as a superior prognostic model in predicting the recurrence of BLCA. The effectiveness of the model was also evaluated in bladder cancer samples via qRT-PCR. Thus, the novel ten-lncRNA signature, constructed based on multi-omics data, had robust prognostic power in predicting the recurrence of BLCA and potential clinical implications as biomarkers.

Integrative Analysis of Next-Generation Sequencing for Next-Generation Cancer Research toward Artificial Intelligence

The rapid improvement of next-generation sequencing (NGS) technologies and their application in large-scale cohorts in cancer research led to common challenges of big data. It opened a new research area incorporating systems biology and machine learning. As large-scale NGS data accumulated, sophisticated data analysis methods became indispensable. In addition, NGS data have been integrated with systems biology to build better predictive models to determine the characteristics of tumors and tumor subtypes. Therefore, various machine learning algorithms were introduced to identify underlying biological mechanisms. In this work, we review novel technologies developed for NGS data analysis, and we describe how these computational methodologies integrate systems biology and omics data. Subsequently, we discuss how deep neural networks outperform other approaches, the potential of graph neural networks (GNN) in systems biology, and the limitations in NGS biomedical research. To reflect on the various challenges and corresponding computational solutions, we will discuss the following three topics: (i) molecular characteristics, (ii) tumor heterogeneity, and (iii) drug discovery. We conclude that machine learning and network-based approaches can add valuable insights and build highly accurate models. However, a well-informed choice of learning algorithm and biological network information is crucial for the success of each specific research question.