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

Abstract

Toxicogenomics combines high throughput molecular technologies with statistical and machine learning approaches to discover a similar group of doses of chemical compounds (DCCs) and genes to explore toxicogenomic biomarkers and their regulatory DCCs. This is also very important in the toxicity study of environmental stressors, synthetic chemicals and drug discovery and development process. Different clustering algorithms are concerned with the discovering of interesting clusters/groups of row or column entities of a dataset. Among those hierarchical clustering (HC) and logistic probabilistic hidden variable model (LPHVM) can identify toxicogenomic biomarkers and their regulatory DCCs forming co-cluster. However, the HC method is very sensitive to outlying observations. On the other hand, though LPHVM is a robust approach, it consumes more time for calculation since it is Expectation-Maximization (EM) based iterative approach. Additionally, the LPHVM creates artificiality problem taking absolute value of the data matrix. Therefore, to overcome these problems in this paper, we proposed a robust hierarchical co-clustering (RHCOC) algorithm to co-cluster genes and DCCs simultaneously with a view to explore toxicogenomic biomarkers and their regulatory DCCs. The performance of the proposed RHCOC algorithm over the github (https://github.com/mdbahadur/rhcoclust).

Multivariate Analysis of Morpho-Physiological Traits Reveals Differential Drought Tolerance Potential of Bread Wheat Genotypes at the Seedling Stage

“…We have used a comparatively newer R package rhcoclust for robust hierarchical co-clustering of our data. This algorithm showed far less error rate than other contemporary clustering algorithms when outlying observations present in the dataset [ 40 ]. It can also be used to create different trait-genotype clustering matrix that can allow researchers to select genotypes of their desired trait groups.…”

“…The extracted clusters were distinct from each other, whereas the genotypes within each cluster were broadly similar to each other. The STIs of the studied traits were normalized and adapted to the library rhcoclust to generate robust hierarchical co-clusters and cluster heatmap [ 40 ]. Prior to cluster analysis, the number of clusters was determined using gap statistic algorithm in fviz_nbclust function of factoextra.…”

“…However, multivariate analysis techniques can be used to explore relationships, classification and parameter prediction within complex data sets as the conclusions are more realistic, meaningful and accurate [ 39 ]. Among the multivariate techniques, the robust hierarchical co-cluster (RHCOC) approach produces a far lower clustering error rate than the conventional hierarchical clustering approaches in presence of outlying observations in the dataset [ 40 ]. PCA-biplot is one of the most effective multivariate analyses to evaluate the traits interaction and genotypic performance and extensively used to dissect the traits correlation in different crop plants [ 41 ].…”

A Comparison of Trend Models for Predicting Tea Production in Bangladesh

“…Therefore, treatment on the outlier observations is very important. There are three ways to obtain robust estimates in presence of outlier observations as deleting the observations with outlier from the dataset, applying the robust methods, and applying conventional methods on the modified dataset (Hasan, Rana, Begum, Rahman and Mollah, 2018a;Hasan, Rana, Begum, Rahman and Mollah, 2018b;Hasan, Badsha & Mollah, 2020). Among these first one is cooperatively easy to use.…”