MGMG: Cell Morphology-Guided Molecule Generation for Drug Discovery

Area of Science:

• Computational chemistry and cheminformatics
• Drug discovery and development
• Phenotypic screening and systems biology

Background:

• Designing novel bioactive molecules is a key challenge in drug discovery.
• Traditional target-based drug discovery is limited for diseases without known targets or reference compounds.
• Existing methods often fail when target information is unavailable or molecular descriptions are insufficient.

Purpose of the Study:

• To introduce Morphology-Guided Molecule Generation (MGMG), a novel approach for phenotypic drug discovery.
• To enable target-agnostic molecule design by integrating cellular morphology and molecular text data.
• To demonstrate MGMG's applicability to both compound and genetic perturbation-based activator design.

Main Methods:

• MGMG integrates cellular morphological profiles from compound treatments with molecular textual descriptions.
• The approach does not require prior knowledge of molecular targets or reference compounds.
• It leverages complementary structural and bioactivity information from both data types.

Main Results:

• MGMG significantly enhances molecule generation performance, especially with limited textual or morphological data.
• The method successfully designs activators from gene overexpression morphology without needing reference compound structures.
• In silico docking shows MGMG-generated molecules have comparable binding affinities to reference compounds, with preserved interactions and structural diversity.

Conclusions:

• MGMG offers a powerful, target-agnostic strategy for designing diverse, bioactivity-aware molecules.
• The approach effectively combines morphological and textual data for guided molecule generation.
• MGMG expands the scope of drug discovery to diseases with unknown targets.