A survey on single and multi omics data mining methods in cancer data classification

Momeni, Zahra; Hassanzadeh, Esmail; Abadeh, Mohammad Saniee; Bellazzi, Riccardo

doi:10.1016/j.jbi.2020.103466

Cited by 35 publications

(20 citation statements)

References 119 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…It is an extremely rare case that a single variable selector based on a filter, wrapper, or embedded technique alone is constructed. The earlier‐mentioned variable selection techniques are usually integrated to generate multiple variable selectors, by ensemble or hybrid methods (Alhenawi et al., 2022; Momeni et al., 2020; Yang et al., 2021). The ensemble methods initiate multiple variable selectors in parallel and combine the outcomes of those variable selectors.…”

Section: General Workflowmentioning

confidence: 99%

Untargeted metabolomics by liquid chromatography‐mass spectrometry for food authentication: A review

Zhong

Wei

et al. 2022

Comp Rev Food Sci Food Safe

View full text Add to dashboard Cite

Food fraud is currently a growing global concern with far-reaching consequences. Food authenticity attributes, including biological identity, geographical origin, agricultural production, and processing technology, are susceptible to food fraud. Metabolic markers and their corresponding authentication methods are considered as a promising choice for food authentication. However, few metabolic markers were available to develop robust analytical methods for food authentication in routine control. Untargeted metabolomics by liquid chromatography-mass spectrometry (LC-MS) is increasingly used to discover metabolic markers. This review summarizes the general workflow, recent applications, advantages, advances, limitations, and future needs of untargeted metabolomics by LC-MS for identifying metabolic markers in food authentication. In conclusion, untargeted metabolomics by LC-MS shows great efficiency to discover the metabolic markers for the authenticity assessment of biological identity, geographical origin, agricultural production, processing technology, freshness, cause of animals' death, and so on, through three main steps, namely, data acquisition, biomarker discovery, and biomarker validation. The applica-

show abstract

Section: General Workflowmentioning

confidence: 99%

Untargeted metabolomics by liquid chromatography‐mass spectrometry for food authentication: A review

Zhong

Wei

et al. 2022

Comp Rev Food Sci Food Safe

View full text Add to dashboard Cite

show abstract

“…For this reason several reviews concentrate solely on them and introduce taxonomies that distinguish, e.g. 'model-agnostic' versus 'model-dependent' methods [28], or exploit an 'earlyintermediate-late' taxonomy [27,[29][30][31][32][33] (described in detail in Appendix A).…”

Section: Introductionmentioning

confidence: 99%

“…techniques [34] or supervised multi-omics prediction models [29,33,35], or survey data-fusion techniques that are either applied to multi-omics data [16,[25][26][27]36], or that apply specific data-fusion techniques (e.g. integrative Bayesian models [13,37] or multimodal neural networks [38]).…”

Section: Introductionmentioning

confidence: 99%

Heterogeneous data integration methods for patient similarity networks

Gliozzo

Mesiti

Notaro

et al. 2022

Briefings in Bioinformatics

View full text Add to dashboard Cite

Patient similarity networks (PSNs), where patients are represented as nodes and their similarities as weighted edges, are being increasingly used in clinical research. These networks provide an insightful summary of the relationships among patients and can be exploited by inductive or transductive learning algorithms for the prediction of patient outcome, phenotype and disease risk. PSNs can also be easily visualized, thus offering a natural way to inspect complex heterogeneous patient data and providing some level of explainability of the predictions obtained by machine learning algorithms. The advent of high-throughput technologies, enabling us to acquire high-dimensional views of the same patients (e.g. omics data, laboratory data, imaging data), calls for the development of data fusion techniques for PSNs in order to leverage this rich heterogeneous information. In this article, we review existing methods for integrating multiple biomedical data views to construct PSNs, together with the different patient similarity measures that have been proposed. We also review methods that have appeared in the machine learning literature but have not yet been applied to PSNs, thus providing a resource to navigate the vast machine learning literature existing on this topic. In particular, we focus on methods that could be used to integrate very heterogeneous datasets, including multi-omics data as well as data derived from clinical information and medical imaging.

show abstract

“…7,8 Therefore, to obtain stable and accurate results as well as better understand the disease process, the direction of study on exploring the association between genes and prognosis has switched from singleomics data analysis to multi-omics datasets. 9 Recently, studies have indicated that integrated analyses of DNA methylation and gene expression could better reveal the regulatory function of DNA methylation and effectively predict the prognosis of patients. [10][11][12] So far, studies on the development of the prognostic model for CRC by integrating DNA methylation and gene expression profiles have been reported.…”

Section: Introductionmentioning

confidence: 99%

“…They often ignored the complexity of the landscape of molecular phenomena underlying CRC 7,8 . Therefore, to obtain stable and accurate results as well as better understand the disease process, the direction of study on exploring the association between genes and prognosis has switched from single‐omics data analysis to multi‐omics datasets 9 . Recently, studies have indicated that integrated analyses of DNA methylation and gene expression could better reveal the regulatory function of DNA methylation and effectively predict the prognosis of patients 10‐12 .…”

Section: Introductionmentioning

confidence: 99%

Development and validation of 3‐CpG methylation prognostic signature based on different survival indicators for colorectal cancer

Huang

Zhang

et al. 2021

Molecular Carcinogenesis

View full text Add to dashboard Cite

Abnormal DNA methylation is considered a vital hallmark to regulate gene expression and influence the development and progression of colorectal cancer (CRC). Although CRC‐related methylation prognostic models have been developed, their clinical application is limited due to the lack of external validation and extension to other survival evaluation indicators. Therefore, this study aimed to develop and validate novel methylation prognostic models correlated with different survival indicators for individualized prognosis prediction for CRC patients. The prognostic‐related CpG sites of methylation‐driven genes screened by the MethylMix algorithm were identified and validated in The Cancer Genome Atlas (TCGA) CRC methylation data and our methylation data. The prognostic models correlated with different survival evaluation indicators (overall survival [OS] and disease‐free survival [DFS]) were developed and validated in the TCGA CRC dataset (N = 376) and our independent CRC dataset (N = 227). We utilized the combination of selected 3‐CpG methylation sites in three genes (DAPP1, FAM3D, and PIGR) to construct a prognostic risk‐score model. In the training dataset, Kaplan–Meier survival analysis demonstrated that high‐risk patients had significantly poorer survival than low‐risk patients (pOS = .0014; pDFS < .001). Then, the 3‐CpG methylation signature was successfully validated as an independent predictor in the testing data set (pOS = .016; pDFS = .016). A prognostic nomogram was constructed and validated. Additionally, mediation analysis revealed the direct effect of the methylation signature on CRC prognosis (pOS = 9.149e−06; pDFS = .001). In summary, our study revealed that the 3‐CpG methylation signature might be a potential prognostic indicator to facilitate individualized survival prediction for CRC patients.

show abstract

A survey on single and multi omics data mining methods in cancer data classification

Cited by 35 publications

References 119 publications

Untargeted metabolomics by liquid chromatography‐mass spectrometry for food authentication: A review

Untargeted metabolomics by liquid chromatography‐mass spectrometry for food authentication: A review

Heterogeneous data integration methods for patient similarity networks

Development and validation of 3‐CpG methylation prognostic signature based on different survival indicators for colorectal cancer

Contact Info

Product

Resources

About