Inferential, robust non-negative matrix factorization analysis of microarray data

Fogel, Paul; Young, S. Stanley; Hawkins, Douglas M.; Ledirac, Nathalie

doi:10.1093/bioinformatics/btl550

Cited by 58 publications

(39 citation statements)

References 21 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Fogel et al used trimmed least squares fitting and dimensionality estimation by mixture modeling to create inferential robust NMF (irNMF) [44]. The method also iteratively removes discordant observations, limiting the impact of outliers on the model.…”

Section: Extensions and Alternatives To Matrix Factorization Methodsmentioning

confidence: 99%

Matrix factorization for transcriptional regulatory network inference

Ochs

Fertig

2012

2012 IEEE Symposium on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB)

View full text Add to dashboard Cite

Inference of Transcriptional Regulatory Networks (TRNs) provides insight into the mechanisms driving biological systems, especially mammalian development and disease. Many techniques have been developed for TRN estimation from indirect biochemical measurements. Although successful when initially tested in model organisms, these regulatory models often fail when applied to data from multicellular organisms where multiple regulation and gene reuse increase dramatically. Non-negative matrix factorization techniques were initially introduced to find non-orthogonal patterns in data, making them ideal techniques for inference in cases of multiple regulation. We review these techniques and their application to TRN analysis.

show abstract

Section: Extensions and Alternatives To Matrix Factorization Methodsmentioning

confidence: 99%

Matrix factorization for transcriptional regulatory network inference

Ochs

Fertig

2012

2012 IEEE Symposium on Computational Intelligence in Bioinformatics and Computational Biology (CIBCB)

View full text Add to dashboard Cite

show abstract

“…On one hand, we can use NMF to do feature selection or information compression, and then use other methods to do classification [5]. On the other hand, for a rank k NMF, we can group the samples into k clusters according to their maximum combinatory coefficients [6]- [11]. For example, for the third column vector v (3) (generally represent a sample in sample-variable data) of V in Fig.…”

Section: A Nonnegative Matrix Factorizationmentioning

confidence: 99%

Two-way combinatorial clustering network

Cao

Liu

2010

2010 2nd International Conference on Future Computer and Communication

View full text Add to dashboard Cite

Two trends in clustering (also called unsupervised classification) problem: from one-way to two-way and from tree structure to net structure, are integrated in this paper to a framework of two-way combinatorial clustering network (TWCCN). The theory of directed branch-connected tree (DBCT) is constructed to describe the model of TWCCN, and algorithms based on nonnegative matrix factorization (NMF) called bootstrap NMF are proposed to build TWCCN. We show the method make sense take examples for the clustering of gene expression data and the problem of phylogenetics in bioinformatics.

show abstract

“…A document is grouped into the cluster where it has the largest projection value. NMF has been applied to many areas in computational biology, ranging from gene expression analysis [17], protein sequence recognition [13], class comparison and prediction [9], cross-platform and cross-species characterization [27], function characterization of genes [22], biological network analysis [29], to biomedical informatics [2]. As an increasingly important tool in computational biology for analysis and interpretation [4], NMF has recently gained more and more attention in biomedical fields.…”

Section: Introductionmentioning

confidence: 99%