Universal Nature of Drug Treatment Responses in Drug-Tissue-Wide Model-Animal Experiments Using Tensor Decomposition-Based Unsupervised Feature Extraction

Abstract

Gene expression profiles of tissues treated with drugs have recently been used to infer clinical outcomes. Although this method is often successful from the application point of view, gene expression altered by drugs is rarely analyzed in detail, because of the extremely large number of genes involved. Here, we applied tensor decomposition (TD)-based unsupervised feature extraction (FE) to the gene expression profiles of 24 mouse tissues treated with 15 drugs. TD-based unsupervised FE enabled identification of the common effects of 15 drugs including an interesting universal feature: these drugs affect genes in a gene-group-wide manner and were dependent on three tissue types (neuronal, muscular, and gastroenterological). For each tissue group, TD-based unsupervised FE enabled identification of a few tens to a few hundreds of genes affected by the drug treatment. These genes are distinctly expressed between drug treatments and controls as well as between tissues in individual tissue groups and other tissues. We also validated the assignment of genes to individual tissue groups using multiple enrichment analyses. We conclude that TD-based unsupervised FE is a promising method for integrated analysis of gene expression profiles from multiple tissues treated with multiple drugs in a completely unsupervised manner.

Adapted tensor decomposition and PCA based unsupervised feature extraction select more biologically reasonable differentially expressed genes than conventional methods

“…One might wonder if the proposed methods, TD- and PCA-based unsupervised FE with optimized SD, are applicable to a more complicated set-up. To investigate this point, we checked the case where multiple drugs are applied to mice whose gene expression of multiple tissues are measured, to which we applied TD-based unsupervised FE 34 .…”

“…In the previous study 34 , although we applied TD-based unsupervised FE to gene expression profiles, there existed some problems. First of all, the number of genes selected was too small to have no false negatives.…”

Tensor decomposition- and principal component analysis-based unsupervised feature extraction to select more reasonable differentially expressed genes: Optimization of standard deviation versus state-of-art methods

“…Thus, considering the optimization of SD enables TD-based unsupervised FE to recognize a greater number of biologically reliable genes than the original TD-based unsupervised FE, which did not include the optimization of SD. 8 Comparison of selected genes between TD-based unsupervised FE [34] and optimal SD with multi-organ data sets…”

“…All processes used were exactly the same as those described in the previous study [34]. After getting u ℓi , the SD was optimized as described in the main text.…”

“…First of all, the number of genes selected was too small to have no false negatives. Using the optimized SD, the number of selected genes increased (Table 8; for more details, e.g., the definition of the four gene sets, neurons and testis, muscle, gastrointestine 1 and 2, see the previous study [34]. This topic has not been discussed herein as it is not directly related to the comparison of the performance between the original TD-based unsupervised FE and that with the optimised SD.…”