Mutational scanning connects genetic variants to phenotype, enabling the interrogation of protein functions, interactions and variant pathogenicity. However, current methodologies cannot efficiently engineer customizable sets of diverse genetic variants in endogenous loci across cellular contexts in high throughput. Here, we combine cytosine and adenine base editors and a prime editor to assess the pathogenicity of a broad spectrum of variants in the epithelial growth factor receptor gene (EGFR). Using pooled base editing and prime editing guide RNA libraries, we install tens of thousands of variants spanning the full coding sequence of EGFR in multiple cell lines and assess the role of these variants in tumorigenesis and resistance to tyrosine kinase inhibitors. Our EGFR variant scan identifies important hits, supporting the robustness of the approach and revealing underappreciated routes to EGFR activation and drug response. We anticipate that multimodal precision mutational scanning can be applied broadly to characterize genetic variation in any genetic element of interest at high and single-nucleotide resolution.

突变扫描将遗传变异与表型联系起来，从而能够询问蛋白质功能、相互作用和变异致病性。然而，目前的方法无法在细胞环境中以高通量有效地设计跨细胞环境中内源性基因座的可定制不同遗传变异集。在这里，我们将胞嘧啶和腺嘌呤碱基编辑器与引物编辑器相结合，以评估上皮生长因子受体基因 （EGFR） 中广谱变体的致病性。使用混合碱基编辑和引物编辑向导 RNA 文库，我们在多个细胞系中安装了数以万计的变体，跨越 EGFR 的完整编码序列，并评估这些变体在肿瘤发生和酪氨酸激酶抑制剂耐药性中的作用。我们的 EGFR 变体扫描可识别重要的命中，支持该方法的稳健性，并揭示 EGFR 激活和药物反应的未被充分认识的途径。我们预计多模态精确突变扫描可以广泛应用于以高和单核苷酸分辨率表征任何感兴趣的遗传元件的遗传变异。