In recent years, drug repositioning is globally popular and plays an important role in pharmaceutical industry, which helps to discover new indications for existing drugs. Unlike the huge cost in R&D stage and high failure rates in clinical trials for new drugs, the advantage of drug repositioning is obvious and with which the safety issues and mechanism of existing drugs are well-studied. 

Along with large-scale omics data of drug treatment and genetic perturbation being released, computational biology and bioinformatics methods show a great advantage in integrating the various types of data for drug target identification and drug repositioning studies.  

However, to our knowledge, a high throughput strategy to reposition drugs in human with high precision is still challenging, perhaps due to several obstacles: batch effects, platform differences and different tissue and cell backgrounds make it difficult for integrative analysis; target gene signatures in existing resources are insufficient to cover known drug targets; for example, only ~430 transcription factors (TFs) are included in The Encyclopedia of DNA Elements (ENCODE); and genetic experiments can only be performed in cell lines or animal models, but not human.  

In particular, pattern-matching tools cannot judge whether the similarities or differences stem from tissue-of-origin backgrounds or relevant/true drug-gene interactions (DGIs), while confining the searches to the same tissue or cell type will severely limits the number of drugs, genes or pathways that can be analyzed. 

A research team led by Prof. Jing-Dong Jackie HAN from CAS-MPG Partner Institute for Computational Biology, Shanghai Institute of Nutrition and Health, Chinese Academy of Sciences constructed a new drug repositioning method using non-tissue specific core signatures from cancer transcriptomes. The study proposes drug-gene modes of action inference and drug specificity assessment systems, and predicts 179,004 drug repositioning candidates for 1,938 specific drugs. 

In the study, researchers developed an in silico screen in human in vivo conditions using a reference of single gene mutations’ non-tissue specific “core transcriptome signatures” (CSs) of 8,476 genes generated from the TCGA database. They developed “core-signature Drug-to-Gene” (csD2G) software to scan 3,546 drug treatment profiles against the reference signatures. csD2G significantly outperformed conventional cell line based gene perturbation signatures, L1000 signature based experimental drug-target prediction and existing drug repositioning methods in both coverage and specificity.  

This research entitled “Accurate Drug Reposition through Non-Tissue Specific Core Signatures from Cancer Transcriptomes” was published on Cell Reports on October 9^th, 2018. 

The study was mainly contributed by Dr. XU Chi and assistant researcher AI Daosheng under the supervision of Prof. Jing-Dong Jackie HAN and was supported by grants from China Ministry of Science and Technology, National Natural Science Foundation of China, and Chinese Academy of Sciences. 

Schematic illustration of drug repositioning method and application 
(Image by Prof. HAN’s team) 

Media Contact:
WANG Jin (Ms.)
Shanghai Institute of Nutrition and Health,
Chinese Academy of Sciences
Email: sibssc@sibs.ac.cn