Quantitative Polypharmacology Profiling Based on a Multifingerprint Similarity Predictive Approach.

Area of Science:

• Computational chemistry
• Drug discovery
• Bioinformatics

Background:

• Accurate prediction of ligand-bioactivity is crucial for drug discovery.
• Existing methods may lack the ability to capture complex polypharmacological profiles.
• Efficiently identifying potential drug targets for small molecules remains a challenge.

Purpose of the Study:

• To introduce a novel quantitative ligand-based bioactivity prediction approach.
• To enable comprehensive polypharmacological profiling of small molecules.
• To provide a reliable tool for accurate ligand-target matching.

Main Methods:

• Utilized a multifingerprint similarity search algorithm with 13 distinct molecular fingerprints.
• Calculated Tanimoto similarity values and analyzed variations in measured biological activity (ΔpIC50).
• Applied the method to a large dataset of 4241 protein targets and 418,485 ligands from ChEMBL.

Main Results:

• Demonstrated robustness and predictive potential through extensive internal and external validation studies.
• Comparative studies confirmed the reliability of the approach against existing ligand-target prediction platforms.
• Successfully applied the algorithm in case studies, highlighting its practical utility.

Conclusions:

• The proposed method offers a robust and reliable approach for quantitative bioactivity prediction and polypharmacological profiling.
• The freely available web platform facilitates high-throughput virtual screening and target identification.
• This tool aids in understanding ligand-target interactions and accelerates drug discovery efforts.