High-Performance Gene Expression Module Analysis Tool and Its Application to Chemical Toxicity Data

Abstract

Gene clustering is one of the main themes of data mining approaches in bioinformatics. Although it has the power to analyze gene function, interpretation of the results becomes increasingly difficult when the number of experiments (samples) exceeds hundreds or more. A new type of clustering called "biclustering," where genes and experiments are coclustered in a large-scale of gene expression data, has been extensively studied in the last decade. We have developed "SAMURAI," an original program that detects all the biclusters or "gene modules" whose genes have similar expression patterns to query profile using the ultrafast data mining algorithm called Linear-time Closed itemset Miner (LCM). Using chemical toxicity dataset from J&J rat liver experiments, we compiled an exhaustive dictionary of gene modules by searching datasets of gene modules with each chemical exposure experiment as query. Through the module analysis, we found that our program can detect up/down-regulated gene sets that significantly represent particular GO functions or KEGG pathways, thereby unraveling reactions and mechanisms common to different toxicochemical treatments of hepatocytes.

Deep surveys of transcriptional modules with Massive Associative K-biclustering (MAK)

“…Biclustering methods have emerged as one of the most popular methodologies (10)(11)(12)(13)(14), because they are able to find local, context-specific patterns, are better at identifying signals in large, noisy datasets with overlapping patterns, and can include information from both data axes as well as additional data. In biology, most biclustering methods have been applied to microarray data (10)(11)(12)(14)(15)(16)(17)(18)(19) but also to other data types such as metabolite levels (20), drug interactions (21)(22)(23), RNA multiple sequence alignment (24), phenotype data (25), protein-protein interaction mass spectrometry data (26)(27)(28), and as part of a machine learning pipeline to identify literature associations (29). Examples of impactful discoveries from these algorithms include functional assignments for uncharacterized genes (15), identification of transcriptional modules (30), identification of transcriptional modules with putative transcription factor (TF) binding sites and support in other data types such as protein interactions, pathway membership, phylogenetic profiles, operon associations, and sequence motifs (13,31), breast cancer classification and prognosis (32), identification of associations between transcriptional modules and environments suitable for predicting microbial response to environmental change (33), and large scale biomedical relationships derived from literature (29).…”

CELLPEDIA: a repository for human cell information for cell studies and differentiation analyses

“…The locations of cells or tissues are shown by voxel models of male and female human bodies provided by National Institute of Information and Communications and Technology (NICT) (15). We also provide precalculated gene modules or biclusters extracted from the collected gene expression data described above using our software program called System for Assembling Modules by Ultra Rapid Algorithm on Itemsets (SAMURAI) (16). A gene module consists of a subset of genes and a subset of experiments, and SAMURAI exhaustively extracts gene modules that share common gene expression patterns in both query and gene expression databases (17).…”

The Role of Molecular Biology in the Biomonitoring of Human Exposure to Chemicals

“…The development of software to analyze gene clusters is another powerful technique. SAMURAI is a program which organizes gene clusters from a library and complements the information with a chemical toxicity dataset, allowing for the assessment of the up/down-regulation of gene sets [49]. Another example of integrative bioinformatics is the Chemical Effects in Biological Systems (CEBS) [50], the integration of a DNA microarray, including proteomic and metabolomic studies, with toxicology data and the queries across “omic” platforms that generate new potential biomarkers of exposure (Figure 2).…”