FORMA Therapeutics, Watertown, MA, USA.
Relay Therapeutics, Cambridge, MA, USA.
SLAS Discov. 2020 Jun;25(5):515-522. doi: 10.1177/2472555220908240. Epub 2020 Feb 28.
DNA-encoded library (DEL) technology has become a prominent screening platform in drug discovery owing to the capacity to screen billions or trillions of compounds in a single experiment. Although numerous successes with DEL technology have been reported, we are unaware of a rigorous examination of the many different variables that can influence a screen's success. Herein, we explore the impact of variable sample sequencing depth on the detection of tool compounds with known affinities toward a given target while simultaneously probing the effect of initial compound input. Our sequencing data confirm reports that high-affinity compounds can be discovered directly from a DEL screen, but we demonstrate that a mismatch between selection output and sequencing quantity can obscure useful ligands. Our results highlight the importance of selection coverage in grasping the entire picture of a DEL screen where the signal of a weak or underrepresented ligand may be suppressed by the inherent noise of a selection. These potential missed ligands may be critical to the success or failure of a drug discovery program.
DNA 编码文库 (DEL) 技术已成为药物发现中一种重要的筛选平台，因为它能够在单次实验中筛选数十亿或数万亿种化合物。尽管 DEL 技术已经取得了许多成功，但我们并不知道有严格检查可以影响筛选成功的许多不同变量。在此，我们探讨了可变样本测序深度对检测具有已知亲和力的工具化合物的影响，同时还研究了初始化合物输入的影响。我们的测序数据证实了高亲和力化合物可以直接从 DEL 筛选中发现的报告，但我们也证明了选择输出与测序数量之间的不匹配会掩盖有用的配体。我们的结果强调了选择覆盖率在掌握 DEL 筛选全貌中的重要性，其中弱或代表性不足的配体的信号可能会被选择的固有噪声所抑制。这些潜在的缺失配体可能对药物发现计划的成败至关重要。
