FABIA: factor analysis for bicluster acquisition

Abstract

Motivation: Biclustering of transcriptomic data groups genes and samples simultaneously. It is emerging as a standard tool for extracting knowledge from gene expression measurements. We propose a novel generative approach for biclustering called ‘FABIA: Factor Analysis for Bicluster Acquisition’. FABIA is based on a multiplicative model, which accounts for linear dependencies between gene expression and conditions, and also captures heavy-tailed distributions as observed in real-world transcriptomic data. The generative framework allows to utilize well-founded model selection methods and to apply Bayesian techniques.Results: On 100 simulated datasets with known true, artificially implanted biclusters, FABIA clearly outperformed all 11 competitors. On these datasets, FABIA was able to separate spurious biclusters from true biclusters by ranking biclusters according to their information content. FABIA was tested on three microarray datasets with known subclusters, where it was two times the best and once the second best method among the compared biclustering approaches.Availability: FABIA is available as an R package on Bioconductor (http://www.bioconductor.org). All datasets, results and software are available at http://www.bioinf.jku.at/software/fabia/fabia.htmlContact: hochreit@bioinf.jku.atSupplementary information: Supplementary data are available at Bioinformatics online.

Region-Based Correspondence Between 3D Shapes via Spatially Smooth Biclustering

“…This paper introduces a novel algorithm, where the rationale is to decompose the data matrix into levels, each corresponding to a different bicluster (as in [13,4]), thus allowing to obtain non-exhaustive and possibly overlapping biclusters. In this class of approaches, sparsity plays a crucial role.…”

“…A recent trend is to use matrix factorization tools [4,11,13,19,31,44], most of which relying on the concept of latent block models [11,26,31,34,44]. These approaches simultaneously arrange the rows and columns of a data matrix into groups of similar response patterns, thus yielding biclusters where the rows/columns belong to only one group, and the data matrix is divided into exhaustive and non overlapping biclusters.…”

Bicluster and Pathway Enrichment Analysis of HCV-induced Cirrhosis and Hepatocellular Carcinoma

“…The generative framework allows to utilize well-founded model selection methods and to apply Bayesian techniques (Hochreiter et al, 2010). In this present study, we performed bicluster analysis using fabia to classify the genes in the progression and identified the differentially expressed genes among healthy liver samples, cirrhotic samples and HCC samples.…”

Algoritmos e técnicas de validação em agrupamento de dados multi-representados, agrupamento possibilístico e bi-agrupamento

“…Além disso, em um conjunto de dados transcriptômicos múltiplas vias genéticas (genetic pathways) podem estar ativos em uma condição e um gene pode participar de diferentes caminhos genéticos sob condições distintas. Para detectar essas interações entre genes, os bi-grupos precisam se sobrepor (Hochreiter et al, 2010). O mesmo pode ser dito sobre documentos textuais descritos por conjuntos de palavras (bag of words), nos quais um documento pode pertencer a categorias diferentes dependendo das palavras consideradas.…”

“…No entanto, os ín-dices clássicos de agrupamento de dados não são diretamente aplicáveis em bi-agrupamentos, i.e., em soluções produzidas por algoritmos de bi-agrupamento. Alguns índices de validação externa de bi-agrupamento foram recentemente propostos (Liu e Wang, 2007;Santamaría et al, 2007;Lee et al, 2009;Hochreiter et al, 2010), invariavelmente em trabalhos nos quais a principal contribuição foi a proposição de um novo algoritmo de bi-agrupamento ou a comparação experimental de algoritmos de bi-agrupamento. Como consequência, propriedades importantes que um índice de validação externa deve possuir foram negligenciadas, acarretando em problemas críticos identificados durante este trabalho.…”