[Neural network modeling of multitarget RAGE inhibitory activity]

Area of Science:

• Computational chemistry
• Molecular modeling
• Pharmacology

Background:

• The Receptor for Advanced Glycation Endproducts (RAGE) and its signaling pathway involving Nuclear Factor kappa B (NF-kB) are implicated in various inflammatory diseases.
• Identifying inhibitors that target the RAGE-NF-kB pathway is crucial for therapeutic development.
• Understanding the specific protein targets within this pathway that mediate inhibitory activity is essential for drug design.

Purpose of the Study:

• To construct predictive models for RAGE inhibitory activity based on compound affinity for RAGE-NF-kB pathway targets.
• To identify key protein targets within the RAGE-NF-kB pathway that are most significant for RAGE inhibitory activity.
• To investigate whether known RAGE inhibitors preferentially target AGE receptors or other components of the signaling cascade.

Main Methods:

• Development of artificial neural network models to correlate compound affinity with RAGE inhibitory activity.
• Creation of a validated database of 183 known RAGE inhibitors and their activity levels.
• Analysis of the RAGE-NF-kB signaling pathway to identify key nodes and target proteins (34 identified).
• Compilation of 3D structural models for 22 target proteins and ensemble molecular docking of 183 RAGE inhibitors.
• Classification model construction using artificial multilayer perceptron neural networks and ROC analysis for prognostic ability evaluation.

Main Results:

• Models were constructed to predict RAGE inhibitory activity, achieving up to 90% accuracy for high activity levels via ROC analysis.
• Sensitivity analysis identified key targets within the RAGE-NF-kB pathway.
• The most significant biotargets for high RAGE inhibitory activity were found to be eight signaling kinases and the transcription factor NF-kB1, rather than AGE receptors themselves.

Conclusions:

• Artificial neural networks effectively model the relationship between compound affinity and RAGE inhibitory activity.
• Known compounds with high RAGE inhibitory activity appear to preferentially inhibit signaling kinases within the RAGE-NF-kB pathway.
• This suggests a potential therapeutic strategy focusing on kinase inhibition for RAGE-mediated conditions.