The optimization of running time for a maximum common substructure-based algorithm and its application in drug design

Area of Science:

• Computational chemistry and drug discovery
• Bioinformatics and cheminformatics

Background:

• High-throughput virtual screening is crucial for identifying bioactive compounds in drug discovery.
• Maximum common substructure (MCS)-based algorithms offer a promising approach for virtual screening.
• A key challenge is balancing the efficiency and accuracy of MCS algorithms in practical applications.

Purpose of the Study:

• To optimize the efficiency of MCS-based virtual screening.
• To evaluate the impact of computation running time on screening accuracy.
• To establish optimal running time thresholds for MCS algorithms.

Main Methods:

• Running time evaluation of the MCS algorithm using WHO essential drugs and FDA-approved small-molecule drugs.
• Variation of allocated running time for MCS-based virtual screening.
• Statistical analysis to correlate running time with screening results and compound structure-activity relationships.

Main Results:

• Running time efficiency can be significantly improved by implementing proper running time thresholds.
• Accuracy is maintained even with optimized running times, with 15-30 seconds identified as a suitable threshold range.
• Quantitative analysis confirms the relevance of compound structure similarity to biological activity, supporting MCS applicability.

Conclusions:

• Optimized MCS algorithms can predict the potential biological activity of small molecules with unknown functions.
• The study provides a generalized conclusion considering CPU speed variations.
• The findings highlight the practical applicability of MCS-based methods in drug candidate virtual screening.