Logic-based machine learning predicts how escitalopram attenuates cardiomyocyte hypertrophy.

Area of Science:

• Cardiovascular Research
• Computational Biology
• Pharmacology

Background:

• Cardiomyocyte hypertrophy predicts heart failure.
• AI and high-throughput screening can identify drugs targeting hypertrophy.

Purpose of the Study:

• Develop LogiRx, a mechanistic machine learning method to predict drug-induced pathways.
• Discover and validate antihypertrophic pathways for existing drugs.

Main Methods:

• Applied logic-based mechanistic machine learning (LogiRx) to predict drug pathways.
• Experimentally validated predictions in cell cultures, mouse models, and patient databases.
• Investigated off-target pathways, including serotonin receptor signaling.

Main Results:

• LogiRx predicted antihypertrophic pathways for seven noncardiac drugs.
• Escitalopram (Lexapro) and mifepristone inhibited cardiomyocyte hypertrophy.
• Escitalopram demonstrated efficacy via an off-target serotonin receptor/PI3Kγ pathway, validated in mice and human databases.
• Patients on escitalopram showed reduced cardiac hypertrophy incidence.

Conclusions:

• LogiRx effectively discovers drug pathways and enables drug repurposing for cardiac remodeling.
• Escitalopram shows potential for treating cardiac hypertrophy through an off-target mechanism.