Integrating population-level and cell-based signatures for drug repositioning

Area of Science:

• Pharmacogenomics
• Computational Biology
• Drug Discovery

Background:

• Drug repositioning offers a cost-effective strategy for identifying new therapeutic uses for existing drugs.
• Drugs with strong human genetic evidence show higher success rates in clinical trials and FDA approval.
• Current drug repositioning methods often focus on single genes, limiting the exploration of broader molecular signatures.

Purpose of the Study:

• To introduce a novel framework, Transcriptome-informed Reversal Distance (TReD), for efficient drug repositioning.
• To integrate disease signatures and drug response profiles using a high-dimensional normed space.
• To quantify the therapeutic reversal potential of candidate drugs in disease-specific cell-based screening.

Main Methods:

• Developed the Transcriptome-informed Reversal Distance (TReD) framework.
• Embedded disease signatures and drug response profiles into a high-dimensional normed space.
• Applied TReD to identify candidate drugs for COVID-19, type 2 diabetes, and Alzheimer's disease.

Main Results:

• Identified 36 potential drug candidates for COVID-19, 16 for type 2 diabetes, and 11 for Alzheimer's disease.
• Achieved significant literature support for identified candidates: 69% for COVID-19, 31% for type 2 diabetes, and 64% for Alzheimer's disease.
• Found that seven COVID-19 and three Alzheimer's disease candidates are already in clinical trials.

Conclusions:

• The TReD framework provides a genetics-anchored approach for drug repositioning.
• Integrating population-level signatures with cell-based screening accelerates therapeutic strategy development.
• This framework has the potential to significantly speed up the discovery of new treatments.