OCTAD: an open workspace for virtually screening therapeutics targeting precise cancer patient groups using gene...

Area of Science:

• Oncology
• Bioinformatics
• Pharmacology

Background:

• Precision medicine tailors cancer treatments to molecular and clinical features.
• Identifying novel drug candidates for specific cancer subtypes requires dedicated resources.
• Gene expression profiles of cancer patients and cellular responses to compounds are increasingly available.

Purpose of the Study:

• To develop an accessible platform for virtually screening drug compounds against precise cancer subtypes.
• To facilitate the discovery of drug candidates by matching cancer expression signatures with compound-induced profiles.
• To address the challenges of multiscale procedures in identifying efficacious drug candidates.

Main Methods:

• Developed Open Cancer TherApeutic Discovery (OCTAD), a web portal and R package.
• Utilized a database of 19,127 patient tissue samples and 12,442 compound expression profiles.
• Implemented deep-learning-based reference tissue selection, signature creation, and drug reversal potency scoring.

Main Results:

• OCTAD enables virtual screening of compounds targeting specific cancer patient groups.
• The platform facilitates the prioritization of small molecules with high potency to reverse cancer gene expression signatures.
• The approach proves efficient and cost-effective for identifying potential drug candidates.

Conclusions:

• OCTAD provides an open workspace for computational and experimental scientists.
• The tool simplifies the complex process of identifying targeted cancer therapeutics.
• OCTAD accelerates the preclinical evaluation of drug candidates in precision oncology.