Robust Co-clustering to Discover Toxicogenomic Biomarkers and Their Regulatory Doses of Chemical Compounds Using Logistic Probabilistic Hidden Variable Model

Abstract

Detection of biomarker genes and their regulatory doses of chemical compounds (DCCs) is one of the most important tasks in toxicogenomic studies as well as in drug design and development. There is an online computational platform “Toxygates” to identify biomarker genes and their regulatory DCCs by co-clustering approach. Nevertheless, the algorithm of that platform based on hierarchical clustering (HC) does not share gene-DCC two-way information simultaneously during co-clustering between genes and DCCs. Also it is sensitive to outlying observations. Thus, this platform may produce misleading results in some cases. The probabilistic hidden variable model (PHVM) is a more effective co-clustering approach that share two-way information simultaneously, but it is also sensitive to outlying observations. Therefore, in this paper we have proposed logistic probabilistic hidden variable model (LPHVM) for robust co-clustering between genes and DCCs, since gene expression data are often contaminated by outlying observations. We have investigated the performance of the proposed LPHVM co-clustering approach in a comparison with the conventional PHVM and Toxygates co-clustering approaches using simulated and real life TGP gene expression datasets, respectively. Simulation results show that the proposed method improved the performance over the conventional PHVM in presence of outliers; otherwise, it keeps equal performance. In the case of real life TGP data analysis, three DCCs (glibenclamide-low, perhexilline-low, and hexachlorobenzene-medium) for glutathione metabolism pathway dataset as well as two DCCs (acetaminophen-medium and methapyrilene-low) for PPAR signaling pathway dataset were incorrectly co-clustered by the Toxygates online platform, while only one DCC (hexachlorobenzene-low) for glutathione metabolism pathway was incorrectly co-clustered by the proposed LPHVM approach. Our findings from the real data analysis are also supported by the other findings in the literature.

Robust Hierarchical Co-clustering to Explore Toxicogenomic Biomarkers and Their Regulatory Doses of Chemical Compounds

“…The fold change gene expression (FCGE) data which computed from this experiment using the equations (1) and (2) are used in most of the toxicogenomics studies. Because FCGE dataset directly reflects the treatment effects (Nyström-Persson et al, 2013;Chung et al, 2015;Nyström-Persson et al, 2017;Hasan et al, 2018Hasan et al, , 2019. The fold change gene expression for the ℎ ( = 1, 2, ⋯ ) sample and for single time point can be computed from the gene expression data of this experiment using the equations:…”

“…It received increasing attention in genetics with the rapid advancement of high throughput molecular profiling technologies such as transcriptomics, proteomics and metabolomics (Hamadeh et al, 2002;Waters and Fostel, 2004;Ancizar-Aristizábal et al, 2014). Statistical algorithms or machine learning techniques together with these high throughput technologies can identify toxicogenomic biomarkers as well as asses chemicals'/drugs' safety (Chung et al, 2015;Hasan et al, 2018). The toxicogenomic biomarkers are those genes or proteins or metabolites which become up/down-regulated exposure to chemicals/drugs.…”

“…In toxicogenomic studies and in drug discovery and development process it is very important to know doses of chemical compounds (DCCs)/drugs similarity or groups according to the mechanism of action as well as groups of genes which have a similar expression patterns in response to these DCCs groups (Madeira and Oliveira, 2004;Afshari et al, 2011;Hasan et al, 2018). The clustering algorithms are concerned with the discovery of interesting groups of row or column entities of the data matrix with large heterogeneity among them of a dataset based on the values of these entities.…”

“…Different clustering algorithms are developed for a variety of application domain which has been proven to be very effective (Ward Jr, 1963;MacQueen, 1967;Bezdek, 1981;Kohonen, 1989). Some of these algorithms have been and are presently being implementing to retrieve the latent patterns in gene expression data (Eisen et al, 1998;Tavazoie et al, 1999;Tamayo et al, 1999;Chung et al, 2015;Hasan et al, 2018;Hasan et al, 2019).…”

“…(Berrar et al, 2003;Gottardo et al, 2006;Upton et al, 2009). Though, logistic probabilistic hidden variable model (LPHVM) (Hasan et al, 2018) can cluster as well as co-cluster the toxicogenomic data based on the predetermined number of clusters in absence and presence of outliers in the dataset, but the problem is it consumes more time for computation. Because, LPHVM is an Expectation-Maximization (EM) algorithm (Dempster et al, 1977) based iterative approach.…”