Identifying multi-target drugs for prostate cancer using machine learning-assisted transcriptomic analysis

Area of Science:

• Oncology
• Pharmacology
• Bioinformatics

Background:

• Prostate cancer remains a significant cause of mortality in men, necessitating novel therapeutic strategies.
• Identifying effective treatments for prostate cancer is a critical area of research.

Purpose of the Study:

• To identify potential drug candidates for prostate cancer using transcriptomic data and the Connectivity Map (CMap) database.
• To investigate the anti-cancer properties of identified candidate drugs and their molecular targets.

Main Methods:

• Transcriptomic data analysis and CMap database integration to identify candidate drugs.
• In vitro cell experiments to assess the effect of candidate drugs on prostate cancer cell proliferation.
• Machine learning and dynamic simulation to predict and explore drug targets (ADRB3 and hERG).

Main Results:

• Bisoprolol was identified as a promising candidate drug for prostate cancer treatment.
• Bisoprolol demonstrated an inhibitory effect on prostate cancer cell proliferation in cell experiments.
• ADRB3 and hERG were identified as potential dual targets for bisoprolol in prostate cancer.

Conclusions:

• Bisoprolol exhibits potential as a multi-target drug for prostate cancer, with beta-adrenergic receptor inhibitors being a feasible therapeutic approach.
• The study highlights a novel research strategy for drug discovery by repurposing drugs based on their side effects.
• Further investigation into bisoprolol's pharmacological actions, toxicity, and mechanisms, particularly concerning ADRB3, is warranted for prostate cancer treatment.