DEL2PH4: Predictive 3D Pharmacophores from DNA-Encoded Library Screening Data

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Drug Discovery

Background:

• DNA-encoded library (DEL) screening identifies small-molecule binders from vast compound libraries.
• Significant structure-activity relationship (SAR) data generated by DEL screening remains underutilized.
• Developing methods to extract actionable insights from DEL data is crucial for drug discovery.

Purpose of the Study:

• To introduce DEL2PH4, an automated workflow for converting DEL screening data into 3D pharmacophore models.
• To enable the extraction of actionable 3D SAR information from DEL screening data.
• To support virtual screening, scaffold hopping, and medicinal chemistry optimization.

Main Methods:

• DEL2PH4 integrates statistically enriched compounds with structurally related unenriched analogs as negative examples.
• The workflow constructs 3D pharmacophore models capturing consensus interaction features.
• The method operates independently of compound resynthesis or activity measurements.

Main Results:

• DEL2PH4 demonstrated strong enrichment of positives over decoy molecules in retrospective benchmarking for a MerTK kinase DEL screen.
• The workflow successfully recovered known active compounds from external datasets.
• The generated pharmacophore models showed consistency with experimentally determined X-ray binding modes.

Conclusions:

• DEL2PH4 provides a general strategy for translating DEL screening outputs into interpretable 3D models.
• The workflow effectively extracts actionable 3D SAR information from DEL screening data.
• DEL2PH4 supports key medicinal chemistry optimization processes like virtual screening and scaffold hopping.