Predicting high-throughput screening results with scalable literature-based discovery methods.

Area of Science:

• Pharmacology
• Computational Biology
• Oncology

Background:

• Drug development costs are rising, prompting interest in repurposing existing drugs.
• In silico methods can predict drug efficacy but often lack empirical validation.
• Cell-based screening is common but may not elucidate the biological mechanisms of action.

Purpose of the Study:

• To evaluate the efficacy of a literature-based computational approach, discovery-by-analogy, for drug repurposing.
• To assess the ability of this method to identify active agents against prostate cancer cells.
• To determine if the method provides a biological basis for its predictions.

Main Methods:

• A scalable literature-based approach, discovery-by-analogy, was employed.
• A large library of agents was screened for activity against prostate cancer cell lines.
• The method allowed retrieval of knowledge used for making predictions.

Main Results:

• The discovery-by-analogy method successfully identified a subset of active agents from a large library.
• The approach demonstrated effectiveness in predicting agents with activity against prostate cancer cells.
• The method provided a plausible biological rationale for the predicted activities.

Conclusions:

• Literature-based computational methods like discovery-by-analogy can effectively guide drug repurposing efforts.
• This approach offers a cost-effective strategy for identifying novel therapeutic uses for existing drugs.
• The ability to provide a biological basis enhances the utility of in silico drug discovery methods.