Regularized Generalized Canonical Correlation Analysis: A Framework for Sequential Multiblock Component Methods

Tenenhaus, Michel; Tenenhaus, Arthur; Groenen, Patrick J. F.

doi:10.1007/s11336-017-9573-x

Cited by 89 publications

(67 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We denote by y ∈ {0, 1} n the disease status or target. X and Z are scaled respectively by √ p and √ q to balance for the different number of features [11] and we assume that X, Z, y have been centred and scaled to unit variance.…”

Section: Methodsmentioning

confidence: 99%

Network Regularization in Imaging Genetics Improves Prediction Performances and Model Interpretability on Alzheimer’s Disease

Guigui

Philippe

Gloaguen

et al. 2019

2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019)

View full text Add to dashboard Cite

Imaging-genetics is a growing popular research avenue which aims to find genetic variants associated with quantitative phenotypes that characterize a disease. In this work, we combine structural MRI with genetic data structured by prior knowledge of interactions in a Canonical Correlation Analysis (CCA) model with graph regularization. This results in improved prediction performance and yields a more interpretable model.

show abstract

Section: Methodsmentioning

confidence: 99%

Network Regularization in Imaging Genetics Improves Prediction Performances and Model Interpretability on Alzheimer’s Disease

Guigui

Philippe

Gloaguen

et al. 2019

2019 IEEE 16th International Symposium on Biomedical Imaging (ISBI 2019)

View full text Add to dashboard Cite

show abstract

“…We selected nine jDR approaches representative of each of these main mathematical formulations (Table 1), focusing on methods able to combine more than two omics, implemented in R or Python, and with software readily available and documented. These jDR approaches are iCluster 15 , Integrative NMF (intNMF) 16 , Joint and Individual Variation Explained (JIVE) 17 , Multiple co-inertia analysis (MCIA) 18 , Multi-Omics Factor Analysis (MOFA) 19 , Multi-Study Factor Analysis (MSFA) 20 , Regularized Generalized Canonical Correlation Analysis (RGCCA) 21 , matrix-tri-factorization (scikit-fusion) 22 and tensorial Independent Component Analysis (tICA) 23 .…”

Section: Joint Dimensionality Reduction Approaches and Principlesmentioning

confidence: 99%

“…RGCCA 21 is one of the most widely used generalizations of CCA to multi-omics data. Similarly to MCIA, RGCCA factorizes each omics into omics-specific factors:…”

Section: Regularized Generalized Canonical Correlation Analysis (Rgcca)mentioning

confidence: 99%

Benchmarking joint multi-omics dimensionality reduction approaches for cancer study

Cantini

Zakeri

Hernandez

et al. 2020

Preprint

View full text Add to dashboard Cite

High-dimensional multi-omics data are now standard in biology. They can greatly enhance our understanding of biological systems when effectively integrated. To achieve this multi-omics data 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 performed a systematic evaluation of nine representative jDR methods using three complementary benchmarks. First, we evaluated their performances in retrieving ground-truth sample clustering from simulated multi-omics datasets. Second, we used TCGA cancer data to assess their strengths in predicting survival, clinical annotations and known pathways/biological processes. Finally, we assessed their classification of multi-omics singlecell data.From these in-depth comparisons, we observed that intNMF performs best in clustering, while MCIA offers a consistent and effective behavior across many contexts. The full code of this benchmark is implemented in a Jupyter notebook -multi-omics mix (momix) -to foster reproducibility, and support data producers, users and future developers.

show abstract

“…To compare the performance of HIM in selecting miRNA-mRNA pairs with more traditional approaches such as correlation analysis, we used a regularized generalized canonical correlation analysis (rgCCA), a componentbased approach that aims to study the relationships between several sets of variables [17]. This analysis selected 36,963 miRNA-mRNA pairs (|r| > 0.5 and FDR < 0.05) from GU dataset and 36,962 miRNA-mRNA pairs (|r| > 0.5 and FDR < 0.05) from the TCGA dataset.…”

Section: Correlation Analysis Of Gu and Tcga Datasetsmentioning

confidence: 99%

Identification of miRNA-mRNA associations in hepatocellular carcinoma using hierarchical integrative model

et al. 2020

View full text Add to dashboard Cite

Background: The established role miRNA-mRNA regulation of gene expression has in oncogenesis highlights the importance of integrating miRNA with downstream mRNA targets. These findings call for investigations aimed at identifying disease-associated miRNA-mRNA pairs. Hierarchical integrative models (HIM) offer the opportunity to uncover the relationships between disease and the levels of different molecules measured in multiple omic studies. Methods: The HIM model we formulated for analysis of mRNA-seq and miRNA-seq data can be specified with two levels: (1) a mechanistic submodel relating mRNAs to miRNAs, and (2) a clinical submodel relating disease status to mRNA and miRNA, while accounting for the mechanistic relationships in the first level. Results: mRNA-seq and miRNA-seq data were acquired by analysis of tumor and normal liver tissues from 30 patients with hepatocellular carcinoma (HCC). We analyzed the data using HIM and identified 157 significant miRNA-mRNA pairs in HCC. The majority of these molecules have already been independently identified as being either diagnostic, prognostic, or therapeutic biomarker candidates for HCC. These pairs appear to be involved in processes contributing to the pathogenesis of HCC involving inflammation, regulation of cell cycle, apoptosis, and metabolism. For further evaluation of our method, we analyzed miRNA-seq and mRNA-seq data from TCGA network. While some of the miRNA-mRNA pairs we identified by analyzing both our and TCGA data are previously reported in the literature and overlap in regulation and function, new pairs have been identified that may contribute to the discovery of novel targets. Conclusion: The results strongly support the hypothesis that miRNAs are important regulators of mRNAs in HCC. Furthermore, these results emphasize the biological relevance of studying miRNA-mRNA pairs.

show abstract

Regularized Generalized Canonical Correlation Analysis: A Framework for Sequential Multiblock Component Methods

Cited by 89 publications

References 41 publications

Network Regularization in Imaging Genetics Improves Prediction Performances and Model Interpretability on Alzheimer’s Disease

Network Regularization in Imaging Genetics Improves Prediction Performances and Model Interpretability on Alzheimer’s Disease

Benchmarking joint multi-omics dimensionality reduction approaches for cancer study

Identification of miRNA-mRNA associations in hepatocellular carcinoma using hierarchical integrative model

Contact Info

Product

Resources

About