Historeceptomic Fingerprints for Drug-Like Compounds

Area of Science:

• Pharmacology
• Bioinformatics
• Computational Biology

Background:

• Drugs often interact with multiple molecular targets (polypharmacology), influencing both therapeutic and adverse effects.
• Understanding a drug's complete molecular fingerprint, including target ensemble and tissue expression, is crucial but often incomplete.
• Adverse drug effects can arise from actions in unintended tissues, highlighting the need for tissue-specific analysis.

Purpose of the Study:

• To develop a computational method for identifying a comprehensive ensemble of drug targets and their tissue-specific expression.
• To create a novel drug fingerprinting system called historeceptomics for a holistic in vivo understanding of drug action.
• To enable applications in drug discovery, personalized medicine, and understanding drug-induced adverse effects.

Main Methods:

• Integrated over 300,000 bioactivity values from the ChEMBL database with 4.2 billion molecular docking scores.
• Combined docking scores with human gene expression data for receptors across various tissues.
• Developed a statistical model to generate significant, drug-specific tissue-receptor (historeceptomics) scores.

Main Results:

• Generated multi-dimensional historeceptomic fingerprints for drugs, offering an intuitive representation of their in vivo activity.
• Established a system that weights drug targets by their expression levels in relevant human tissues.
• Created a comprehensive molecular fingerprint that accounts for polypharmacology and tissue distribution.

Conclusions:

• Historeceptomics provides a novel, interpretable, and holistic view of drug mechanisms of action.
• This approach facilitates the identification of molecular signatures for polypharmacologic drugs and prediction of tissue-specific adverse effects.
• The historeceptomics system offers significant potential for advancing drug discovery, personalized medicine, and hypothesis generation.