We are constructing the "Percellome Database" consisting of the transcriptomes of liver and other organs of mice exposed to different types of chemicals. Using this database we are systematically elucidating the molecular mechanism of toxicity for the development of a new toxicity prediction technology.
The database already contains comprehensive transcriptomic data on 160 chemicals. Acute toxicity of a chemical can be predicted to a certain extent by searching the similarity of the transcriptomes obtained by the single-dose exposure experiments.
In addition, we are performing the repeated-dose experiments on the selected chemicals in order to analyze the correlation between the transcriptome and the epigenome i.e. histone modification by ChIP-Seq and genomic DNA methylation by WGBS for the understanding of the molecular mechanism of the chronic toxicity.
Now, we are attempting to expand the scope of analysis with higher efficiency by introducing the artificial intelligence technologies. This approach should maximize the value of the Percellome database and optimize the utility of the epigenome data for the prediction of chronic toxicity from short-term repeated dosing toxicity studies, and for the progress of practical in silico toxicity prediction in the near future.