Spike-and-slab Lasso biclustering

Moran, Gemma E.; Ročková, Veronika; George, Edward I.

doi:10.1214/20-aoas1385

Cited by 13 publications

(6 citation statements)

References 41 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For the problem of biclustering, i.e. identifying clusters using only subsets of their associated features, Moran et al [34] utilized the factor model (7.4) in which both the factors η = [η T 1 , . .…”

Section: Methodological Extensionsmentioning

confidence: 99%

“…, η T n ] ∈ R n×K and the loadings are sparse. To achieve a doubly sparse representation, Moran et al [34] placed an SSL prior coupled with an IBP prior on the factors and an SSL prior coupled with a beta-Bernoulli prior on the loadings. An EM algorithm with a variational step was developed to implement spike-and-slab LASSO biclustering.…”

Section: Methodological Extensionsmentioning

confidence: 99%

See 1 more Smart Citation

Spike-and-Slab Meets LASSO: A Review of the Spike-and-Slab LASSO

Bai,

Rockova,

George

2020

Preprint

Self Cite

View full text Add to dashboard Cite

High-dimensional data sets have become ubiquitous in the past few decades, often with many more covariates than observations. In the frequentist setting, penalized likelihood methods are the most popular approach for variable election and estimation in high-dimensional data. In the Bayesian framework, spike-and-slab methods are commonly used as probabilistic constructs for high-dimensional modeling. Within the context of univariate linear regression, Ročková and George [48] introduced the spike-and-slab LASSO (SSL), an approach based on a prior which forms a continuum between the penalized likelihood LASSO and the Bayesian point-mass spike-and-slab formulations. Since its inception, the spike-and-slab LASSO has been extended to a variety of contexts, including generalized linear models, factor analysis, graphical models, and nonparametric regression. The goal of this paper is to survey the landscape surrounding spike-and-slab LASSO methodology. First we elucidate the attractive properties and the computational tractability of SSL priors in high dimensions. We then review methodological developments of the SSL and outline several theoretical developments. We illustrate the methodology on both simulated and real datasets.

show abstract

Section: Methodological Extensionsmentioning

confidence: 99%

Section: Methodological Extensionsmentioning

confidence: 99%

Spike-and-Slab Meets LASSO: A Review of the Spike-and-Slab LASSO

Bai,

Rockova,

George

2020

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…DNA microarray was used in various areas of research, including gene discovery [ 6 ], the diagnosis of disease [ 7 ], and drug findings [ 8 ]. Microarrays are used to identify the functions of the genes and the mechanisms that underlie diseases.…”

Section: Introductionmentioning

confidence: 99%

Healthcare Biclustering-Based Prediction on Gene Expression Dataset

Ramkumar¹,

Basker²,

Duraisamy³

et al. 2022

BioMed Research International

View full text Add to dashboard Cite

In this paper, we develop a healthcare biclustering model in the field of healthcare to reduce the inconveniences linked to the data clustering on gene expression. The present study uses two separate healthcare biclustering approaches to identify specific gene activity in certain environments and remove the duplication of broad gene information components. Moreover, because of its adequacy in the problem where populations of potential solutions allow exploration of a greater portion of the research area, machine learning or heuristic algorithm has become extensively used for healthcare biclustering in the field of healthcare. The study is evaluated in terms of average match score for nonoverlapping modules, overlapping modules through the influence of noise for constant bicluster and additive bicluster, and the run time. The results show that proposed FCM blustering method has higher average match score, and reduced run time proposed FCM than the existing PSO-SA and fuzzy logic healthcare biclustering methods.

show abstract

“…Like FABIA, BicMix [ 8 ] and Spike-and-Slab Lasso Biclustering (SSLB) [ 23 ] use sparsity-inducing priors. The key difference with BicMix and SSLB is that they allow for both sparse and dense biclusters, and adapt the sparsity constraints to each bicluster.…”

Section: Methodsmentioning

confidence: 99%

Comparison of sparse biclustering algorithms for gene expression datasets

Nicholls

Wallace

2021

Briefings in Bioinformatics

View full text Add to dashboard Cite

Motivation Gene clustering and sample clustering are commonly used to find patterns in gene expression datasets. However, genes may cluster differently in heterogeneous samples (e.g. different tissues or disease states), whilst traditional methods assume that clusters are consistent across samples. Biclustering algorithms aim to solve this issue by performing sample clustering and gene clustering simultaneously. Existing reviews of biclustering algorithms have yet to include a number of more recent algorithms and have based comparisons on simplistic simulated datasets without specific evaluation of biclusters in real datasets, using less robust metrics. Results We compared four classes of sparse biclustering algorithms on a range of simulated and real datasets. All algorithms generally struggled on simulated datasets with a large number of genes or implanted biclusters. We found that Bayesian algorithms with strict sparsity constraints had high accuracy on the simulated datasets and did not require any post-processing, but were considerably slower than other algorithm classes. We found that non-negative matrix factorisation algorithms performed poorly, but could be re-purposed for biclustering through a sparsity-inducing post-processing procedure we introduce; one such algorithm was one of the most highly ranked on real datasets. In a multi-tissue knockout mouse RNA-seq dataset, the algorithms rarely returned clusters containing samples from multiple different tissues, whilst such clusters were identified in a human dataset of more closely related cell types (sorted blood cell subsets). This highlights the need for further thought in the design and analysis of multi-tissue studies to avoid differences between tissues dominating the analysis. Availability Code to run the analysis is available at https://github.com/nichollskc/biclust_comp, including wrappers for each algorithm, implementations of evaluation metrics, and code to simulate datasets and perform pre- and post-processing. The full tables of results are available at https://doi.org/10.5281/zenodo.4581206.

show abstract

Spike-and-slab Lasso biclustering

Cited by 13 publications

References 41 publications

Spike-and-Slab Meets LASSO: A Review of the Spike-and-Slab LASSO

Spike-and-Slab Meets LASSO: A Review of the Spike-and-Slab LASSO

Healthcare Biclustering-Based Prediction on Gene Expression Dataset

Comparison of sparse biclustering algorithms for gene expression datasets

Contact Info

Product

Resources

About