A platform for target prediction of phenotypic screening hit molecules

Area of Science:

• Drug discovery and development
• Computational chemistry
• Infectious diseases

Background:

• Phenotypic drug discovery is crucial for infectious diseases but target identification is difficult.
• Knowing the molecular target and binding pose aids compound optimization and project prioritization.

Purpose of the Study:

• To develop an in silico platform for predicting molecular targets of phenotypic screening hits.
• To provide researchers with 3D structures of predicted targets bound to active molecules.

Main Methods:

• Developed a web-based computational platform integrating ligand and protein structure information.
• Utilized the platform to generate ranked hypotheses for molecular targets.
• Applied the platform to Mycobacterium tuberculosis drug discovery as a case study.

Main Results:

• The platform successfully predicted molecular targets for 60% of tested compounds with known ligand similarity.
• Generated 3D structures of predicted targets complexed with active molecules.

Conclusions:

• The in silico platform offers an efficient method for target hypothesis generation in phenotypic drug discovery.
• This approach can accelerate the identification of drug targets and optimize lead compounds, particularly for tuberculosis.