Pharmacogenomic Drug-Target Network Analysis Reveals Similarity Profiles Among FDA-Approved Cancer Drugs

Area of Science:

• Pharmacogenomics and Computational Biology
• Oncology Drug Discovery
• Systems Biology

Background:

• Identifying precise molecular targets for cancer therapeutics is a major challenge in oncology.
• Many approved anticancer drugs lack complete target profiles, hindering understanding of their mechanisms and applications.
• This study addresses the need for novel, biologically relevant connections between anticancer drugs and protein-coding genes.

Purpose of the Study:

• To establish novel, biologically meaningful relationships between anticancer drugs and protein-coding genes.
• To develop and validate a new drug similarity index (B-index) based on shared gene targets.
• To identify potential new oncology drug targets and therapeutic applications through network analysis.

Main Methods:

• Integrated transcriptomic data with drug activity data from the NCI-60 cancer cell line panel.
• Analyzed interactions between 124 Food and Drug Administration (FDA)-approved anticancer drugs and 399 cancer-related genes.
• Developed the B-index for drug similarity based on shared gene targets and compared it with chemical structural similarity.

Main Results:

• Identified 1304 statistically significant drug-gene relationships, forming a large-scale pharmacogenomic interaction network.
• Clustering based on the B-index grouped drugs with common targets, aligning with known drug classes and structures.
• Validated the B-index using known drug pairs and discovered novel gene associations for potential drug repurposing.

Conclusions:

• Presented a comprehensive network-based strategy for elucidating cancer drug targets using gene expression and drug activity data.
• The B-index offers an alternative to chemical similarity metrics, facilitating the discovery of new therapeutic relationships.
• Findings pave the way for proposing novel oncology drug targets and informing drug repositioning strategies.