International Journal of Molecular Medical Science, 2024, Vol.14, No.5, 274-292 http://medscipublisher.com/index.php/ijmms 278 metabolism. Metabolomic data helps in understanding the metabolic alterations in cancer cells and can provide insights into the metabolic pathways that are dysregulated in cancer. 3.2 Advantages of multi-omics approaches The integration of multi-omics data offers several advantages in cancer research. The Cancer Genome Atlas (TCGA) provides sequencing data for various cancers from different platforms, including gene expression, DNA methylation, and copy number data (Network, 2012; Qiu et al., 2020; Qiu et al., 2021a; Qiu et al., 2021b). These different types of data alone cannot fully describe the molecular mechanisms of cancer, but they complement each other and cover highly organized molecular and cellular events. Some cancer subtype prediction models integrate different omics data to capture the complexity of phenotypes and the heterogeneity of biological processes (Wang et al., 2014; Ritchie et al., 2015). Compared to models using single omics data (such as gene expression), models utilizing multi-omics data offer a more comprehensive understanding of the molecular mechanisms underlying specific biological processes or complex diseases (de Hijas-Liste et al., 2014). By combining different types of omics data, researchers can gain a more comprehensive understanding of the molecular mechanisms underlying cancer. This holistic approach allows for the identification of novel biomarkers and therapeutic targets that may not be apparent when analyzing a single type of omics data. Multi-omics approaches also improve the accuracy of cancer prognosis and diagnosis. For instance, integrating genomics, transcriptomics, and epigenomics data can enhance the prediction of cancer outcomes and the identification of cancer subtypes (Kim et al., 2014). Additionally, multi-omics data integration can reveal complex interactions between different molecular layers, providing deeper insights into the pathogenesis of cancer (Nicora et al., 2020). At the same time, advancements in technology and decreasing costs have enabled international collaborations such as the International Cancer Genome Consortium and The Cancer Genome Atlas (TCGA) to perform multi-platform sequencing of thousands of tumors, facilitating the transition to integrated multi-omics cancer research (Hudson et al., 2010; Tomczak et al., 2015). Additionally, large projects such as GTEx (Consortium, 2020), ENCODE (Consortium, 2012) , ROADMAP (Kundaje et al., 2015) , and the Human Cell Atlas (Regev et al., 2017) have publicly released whole-genome and tissue-specific molecular maps. These large, publicly available datasets allow researchers to study disease-related tissues across various biological and multi-omics layers (Hasin et al., 2017) and provide deep insights into the connections between risk factors and diseases. 3.3 Challenges in multi-omics data integration Despite its advantages, multi-omics data integration poses several challenges. One major challenge is the heterogeneity of the data, as different types of omics data have different characteristics and scales. Integrating these diverse datasets requires sophisticated computational methods and tools. Additionally, as analytical methods continue to evolve, the approaches for integrating multi-omics datasets are becoming increasingly diverse. New multi-omics integration tools are continuously being developed, making the task of selecting the most suitable integration tool from the numerous available options both complex and time-consuming. Another challenge is the high dimensionality of omics data, which can lead to issues with data sparsity and noise. Effective data integration methods must be able to handle these issues to extract meaningful insights (Silverbush et al., 2019). Additionally, the uneven maturity of different omics technologies can hinder the translation of multi-omics approaches into clinical practice. Due to the uneven maturity of different omics technologies and the high dimensionality of omics data, integrated multi-omics datasets may not always have significant research value. Finally, the complexity of biological systems means that multi-omics data integration must account for intricate interactions between different molecular layers : from genomics and epigenomics to transcriptomics, proteomics, and metabolomics, and back to genomics and epigenomics. This requires advanced modeling techniques and a deep understanding of the underlying biology (Wang et al., 2016). In conclusion, while multi-omics data integration holds great promise for advancing cancer research, it also presents significant challenges that must be addressed to fully realize its potential.
RkJQdWJsaXNoZXIy MjQ4ODYzNQ==