HIDEEP: a systems approach to predict hormone impacts on drug efficacy based on effect paths

Area of Science:

• Computational biology and bioinformatics
• Pharmacology and drug discovery
• Endocrinology and molecular medicine

Background:

• Human endogenous hormones significantly influence drug efficacy.
• Hormone status variability necessitates understanding hormone-drug interactions for precision medicine.
• Current understanding of these molecular interplay mechanisms remains limited.

Purpose of the Study:

• To develop an in silico method for predicting interactions between human endogenous hormones and drugs.
• To identify potential impacts of 283 hormones on 590 drugs across 20 diverse diseases.
• To elucidate molecular mechanisms underlying hormone-drug interplay.

Main Methods:

• Extraction of hormone and drug effect pathways from a large-scale molecular network.
• Analysis of molecular network data including protein interactions, transcriptional regulations, and signaling interactions.
• Prediction of hormone-drug interactions based on the close intersection of their respective effect paths.

Main Results:

• Successfully developed and validated an in silico method for predicting hormone-drug interactions.
• The method demonstrated high accuracy in distinguishing known hormone-drug pairs from random pairs in blind experiments.
• Identified potential interactions for 283 hormones and 590 drugs across various disease contexts.

Conclusions:

• The developed in silico method is effective for predicting significant hormone-drug interactions.
• This approach aids in understanding the molecular basis of how hormones modulate drug efficacy.
• Findings support the advancement of precision medicine by providing insights into personalized therapeutic strategies.