Wisdom of crowds for synthetic accessibility evaluation

Area of Science:

• Medicinal Chemistry
• Computational Chemistry
• Drug Discovery

Background:

• Synthetic accessibility evaluation is crucial for drug discovery.
• Current methods rely on medicinal chemists, leading to low throughput.
• A rapid assessment for vast numbers of compounds is needed.

Purpose of the Study:

• To investigate the feasibility of using crowdsourcing for synthetic accessibility evaluation.
• To assess the accuracy of synthetic accessibility scores derived from semi-experts.
• To address the throughput limitations in evaluating large compound libraries.

Main Methods:

• Utilized crowdsourcing by incorporating semi-experts.
• Applied a statistical aggregation method to combine judgments.
• Conducted experiments to validate the approach.

Main Results:

• Accurate synthetic accessibility scores were obtained.
• Statistical aggregation of semi-expert judgments proved effective.
• Crowdsourcing offers a viable solution for high-throughput evaluation.

Conclusions:

• Crowdsourcing with semi-experts and statistical aggregation is a feasible and accurate method for synthetic accessibility evaluation.
• This approach can significantly enhance throughput in drug discovery.
• Overcomes the bottleneck of relying solely on expert medicinal chemists.