JOINT AND INDIVIDUAL VARIATION EXPLAINED (JIVE) FOR INTEGRATED ANALYSIS OF MULTIPLE DATA TYPES

By Eric F. Lock, Katherine A. Hoadley, James Stephen Marron, Andrew B. Nobel. Annals of Applied Statistics, 2011

536 citations recorded by the source corpus.

Research in several fields now requires the analysis of datasets in which multiple high-dimensional types of data are available for a common set of objects. In particular, The Cancer Genome Atlas (TCGA) includes data from several diverse genomic technologies on the same cancerous tumor samples. In this paper we introduce Joint and Individual Variation Explained (JIVE), a general decomposition of variation for the integrated analysis of such datasets. The decomposition consists of three terms: a low-rank approximation capturing joint variation across data types, low-rank approximations for structured variation individual to each data type, and residual noise. JIVE quantifies the amount of joint variation between data types, reduces the dimensionality of the data, and provides new directions for the visual exploration of joint and individual structure. The proposed method represents an extension of Principal Component Analysis and has clear advantages over popular two-block methods such as Canonical Correlation Analysis and Partial Least Squares. A JIVE analysis of gene expression and miRNA data on Glioblastoma Multiforme tumor samples reveals gene-miRNA associations and provides better characterization of tumor types.

Subfields: Molecular Biology

View Paper: https://www.semanticscholar.org/paper/736bc4c904845ac3bece9f6db10d0d4d36298d3f

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Molecular Biology

Citations in the corpus

Computational principles and challenges in single-cell data integration (2021, methodology): “These approaches include unsupervised dimensionality reduction methods that represent different flavors of matrix factorization, such as canonical correlation analysis (CCA, implemented in Seurat 18,22 ), MOFA 69 , JIVE 70 , PLS 71 , MCIA 72 and iNMF 62 , among others.”
More Is Better: Recent Progress in Multi-Omics Data Integration Methods (2017, background): “In other words, JIVE factors the original data input matrix (gene expression etc.) into two lower ranked representative portions W c (shared factor) and W s (data-speciﬁc factor), dependent on H c and H s (Lock et al., 2013).”
A review on machine learning principles for multi-view biological data integration (2016, methodology): “Although not in a non-negative setting but closely relevant, joint and individual clustering (JIC) [117] and joint and individual variation explained (JIVE) [118] use both common and view-specific coefficient matrices.”
Multi-omics Data Integration, Interpretation, and Its Application (2020, methodology): “54,55 Joint and individual structure corresponds to r and r i dimensional subspace that explains variation across multiple data sets and within data sets, respectively.”
Learning to Separate Object Sounds by Watching Unlabeled Video (2018, methodology): “We compare to several existing methods [43,50,56,65] and multiple baselines:”
Machine Learning and Integrative Analysis of Biomedical Big Data (2019, background): “JIVE is an integrative EDA method that decomposes a dataset into a sum of three terms: two low-rank approximation terms, one for capturing joint structure across data types and other for capturing structure individual to each data, and a term for residual noise.”
Joint probabilistic modeling of single-cell multi-omic data with totalVI (2021, background): “Like other multi-omics analysis methods [31, 50, 51], totalVI assumes that RNA and protein measurements are generated from the same latent space of cells that captures their state.”
Co-Separating Sounds of Visual Objects (2019, methodology): “• AV-Loc [34], JIVE [28], Sparse CCA [25]: We use results reported in [12] to compare to these methods for the audio denoising benchmark AV-Bench.”
Integration strategies of multi-omics data for machine learning analysis (2021, methodology): “JIVE/aJIVE [151,152] Matrix factorization Disease subtyping, systemic knowledge, module detection JIVE was used as a dimension reduction technique to improve survival prediction of patients with glioblastoma from mRNA, miRNA and methylation data [153].”