Chemical genomics data can be explored to identify networks of genes that may collectively define a disease phenotype and/or predict new treatments based on the identified connections. We have applied a chemocentric informatics protocol to predict novel chemical-gene-disease connections that could potentially aid in the prioritization of pharmacotherapy for autism. Four different gene signatures obtained from post-mortem human brain cells and peripheral cell lines were used to query the Connectivity Map (cmap, http://www.broad.mit.edu/cmap/) to identify common drugs that could reverse the autism gene signatures. Further analysis based on structural identity comparisons led to a list of 10 common hits that belonged to four major pharmacologic classes: androgens, anticholinergics, ion channel blockers, and vitamins. Mining the biomedical literature as well as specialized network pharmacology databases identified additional strong connections between our hit compounds and autism suggesting that these drugs could be reprofiled to treat the disease.

Chemocentric informatics analysis of "omics" data identifies novel chemical-gene-disease associations in autism

247th National Meeting (2014)

The official archive and repository for all ACS Meetings & Expositions abstracts and select posters or presentations.