A data-driven journey using results from target-based drug discovery for target deconvolution in phenotypic...

Area of Science:

• Drug Discovery and Development
• Medicinal Chemistry
• Pharmacology

Background:

• Drug discovery employs target-based and phenotypic screening methods, each with distinct advantages and limitations.
• Phenotypic screening identifies compounds with desired effects but often lacks clear target identification, hindering optimization.
• Target deconvolution, crucial for understanding drug mechanisms, relies on highly selective tool compounds.

Purpose of the Study:

• To develop an automated method for identifying highly selective ligands from large bioactivity databases.
• To utilize these selective ligands for target deconvolution in phenotypic screening.
• To discover novel anti-cancer mechanisms through phenotypic screening and target identification.

Main Methods:

• Mining the ChEMBL database (over 20 million bioactivity data points) to identify selective ligands.
• Developing and applying an automated method for selecting high-selectivity compounds.
• Purchasing 87 selected compounds and screening them against 60 cancer cell lines for phenotypic effects.

Main Results:

• A novel automated method for selecting highly selective ligands was successfully developed.
• Screening identified several compounds exhibiting selective inhibition of cancer cell growth.
• Combined phenotypic data and ChEMBL bioactivity revealed potential novel anti-cancer mechanisms.

Conclusions:

• Automated selection of highly selective compounds is effective for target deconvolution.
• These selective compounds can guide phenotypic screening towards identifying underlying drug targets.
• The study suggests novel therapeutic strategies for anti-cancer drug discovery.