Functional genomic screening with CRISPR has provided a powerful and precise new way to interrogate the phenotypic consequences of gene manipulation in high-throughput, unbiased analyses. However, some experimental paradigms prove especially challenging and require carefully and appropriately adapted screening approaches. In particular, negative selection (or sensitivity) screening, often the most experimentally desirable modality of screening, has remained a challenge in drug discovery. Here we assess whether our new, modular genome-wide pooled CRISPR library can improve negative selection CRISPR screening and add utility throughout the drug development pipeline. Our pooled library is split into three parts, allowing it to be scaled to accommodate the experimental challenges encountered during drug development, such as target identification using unlimited cell numbers compared with target identification studies for cell populations where cell numbers are limiting. To test our new library, we chose to look for drug–gene interactions using a well-described small molecule inhibitor targeting poly(ADP-ribose) polymerase 1 (PARP1), and in particular to identify genes which sensitise cells to this drug. We simulate hit identification and performance using each library partition and support these findings through orthogonal drug combination cell panel screening. We also compare our data with a recently published CRISPR sensitivity dataset obtained using the same PARP1 inhibitor. Overall, our data indicate that generating a comprehensive CRISPR knockout screening library where the number of guides can be scaled to suit the biological question being addressed allows a library to have multiple uses throughout the drug development pipeline, and that initial validation of hits can be achieved through high-throughput cell panels screens where clinical grade chemical or biological matter exist.

中文翻译：

---

用于药物开发筛选的灵活、汇集的 CRISPR 库

使用 CRISPR 进行功能基因组筛选提供了一种强大而精确的新方法，可以在高通量、无偏见的分析中询问基因操作的表型结果。然而，一些实验范式证明特别具有挑战性，需要仔细和适当调整筛选方法。特别是，负选择（或敏感​​性）筛选通常是实验上最理想的筛选方式，仍然是药物发现中的一个挑战。在这里，我们评估我们新的、模块化的全基因组汇集 CRISPR 文库是否可以改善负选择 CRISPR 筛选并在整个药物开发管道中增加实用性。我们的合并库分为三个部分，可以扩展以适应药物开发过程中遇到的实验挑战，例如，与细胞数量有限的细胞群的目标识别研究相比，使用无限细胞数进行目标识别。为了测试我们的新文库，我们选择使用描述良好的靶向聚（ADP-核糖）聚合酶 1 (PARP1) 的小分子抑制剂来寻找药物-基因相互作用，特别是确定使细胞对这种药物敏感的基因。我们使用每个库分区模拟命中识别和性能，并通过正交药物组合细胞面板筛选支持这些发现。我们还将我们的数据与最近发布的使用相同 PARP1 抑制剂获得的 CRISPR 敏感性数据集进行了比较。总体，