FL-DTD: an integrated pipeline to predict the drug interacting targets by feedback loop-based network analysis

Area of Science:

• Computational Biology
• Pharmacology
• Bioinformatics

Background:

• Drug target discovery is crucial for understanding drug mechanisms of action (MoA), including therapeutic effects and side effects.
• Current methods are often labor-intensive and struggle to identify tissue-specific targets, particularly those with low drug binding affinity.

Purpose of the Study:

• To propose an integrated pipeline, FL-DTD (Feedback Loop-based Drug Target Discovery), for predicting tissue-specific drug-interacting targets of novel compounds.
• To address limitations of existing methods by enabling identification of tissue-specific targets and those with weaker binding affinities.

Main Methods:

• FL-DTD analyzes network responses to drug perturbation, based on the hypothesis that cells in homeostasis activate feedback loops in response to drugs.
• The method was evaluated using expression data from estrogen stimulation, gene manipulation, and drug perturbation experiments.
• Applied to 500 natural compounds using STAT3 as a target protein, predicting compounds with STAT3-interacting activities.

Main Results:

• FL-DTD demonstrated good performance in identifying annotated drug targets during evaluation.
• The pipeline successfully predicted five compounds with STAT3-interacting activities from a dataset of 500 natural compounds.
• Experimental validation confirmed the STAT3-interacting activities of four of the five predicted compounds.

Conclusions:

• FL-DTD accurately predicts drug-interacting targets, offering a valuable tool for drug target discovery.
• The tissue-specific nature of FL-DTD enhances its utility for identifying relevant targets in specific biological contexts.
• The method's ability to predict targets with weaker binding affinities expands its applicability in drug discovery research.