Chemical structural novelty: on-targets and off-targets

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Pharmacology

Background:

• Quantitative comparison of drug structures is crucial for drug discovery and development.
• Both 2D topological and 3D binding-related characteristics inform drug similarity.
• Understanding chemical similarity-pharmacological novelty is key to predicting drug effects.

Purpose of the Study:

• To develop and apply a framework for combining multiple drug similarity computations.
• To assess the utility of 2D and 3D similarity in predicting drug targets, including off-targets.
• To investigate the relationship between chemical similarity and pharmacological novelty.

Main Methods:

• A novel framework was developed to integrate 2D and 3D similarity metrics.
• The framework was applied to a dataset of 358 drugs with overlapping pharmacology.
• A single similarity score was generated for new molecules against known sets, combining 2D and 3D data.

Main Results:

• The combined 2D/3D similarity approach produced a single predictive score.
• 3D similarity offered a significant advantage over 2D similarity for predicting off-targets.
• High 3D similarity with low 2D similarity indicated novel scaffolds with distinct target modulation.

Conclusions:

• The developed framework effectively combines 2D and 3D drug similarity for quantitative comparison.
• 3D structural information is particularly valuable for predicting unintended drug interactions (off-targets).
• Novel chemical scaffolds with high 3D but low 2D similarity are associated with unique pharmacological profiles.