免疫检查点抑制剂（ICI）已经彻底改变了多种癌症的治疗方法。然而，大多数患者都会遇到耐药性的问题。基因之间的合成活力（SV）可能会诱导耐药性。在本研究中，我们建立了SV标记来预测ICI治疗黑色素瘤的疗效。我们通过随机森林分类器收集了特征并预测了SV基因对。本研究基于CRISPR/Cas9筛选结果确定了SV基因对标记以预测黑色素瘤患者对ICI的反应。本研究根据14个特征预测了可靠的SV基因对。经过CRISPR/Cas9筛选，我们鉴定了1,861对与多种癌症类型的预后相关的SV基因对。接下来，我们构建了六个SV基因对标记，以预测黑色素瘤患者对ICI的耐药性。该研究将六个SV基因对标记应用于将黑色素瘤患者分为高风险组和低风险组。高风险黑色素瘤患者在ICI治疗后的反应更差。免疫景观分析显示，高风险黑色素瘤患者的自然杀伤细胞和CD8+ T细胞浸润程度较低。总而言之，这个14个特征分类器准确预测了癌症的可靠SV基因对。六个SV基因对标记可以预测ICI的耐药性。© 2023. 作者(们)。

Immune checkpoint inhibitors (ICI) have revolutionized the treatment for multiple cancers. However, most of patients encounter resistance. Synthetic viability (SV) between genes could induce resistance. In this study, we established SV signature to predict the efficacy of ICI treatment for melanoma.We collected features and predicted SV gene pairs by random forest classifier. This work prioritized SV gene pairs based on CRISPR/Cas9 screens. SV gene pairs signature were constructed to predict the response to ICI for melanoma patients.This study predicted robust SV gene pairs based on 14 features. Filtered by CRISPR/Cas9 screens, we identified 1,861 SV gene pairs, which were also related with prognosis across multiple cancer types. Next, we constructed the six SV pairs signature to predict resistance to ICI for melanoma patients. This study applied the six SV pairs signature to divide melanoma patients into high-risk and low-risk. High-risk melanoma patients were associated with worse response after ICI treatment. Immune landscape analysis revealed that high-risk melanoma patients had lower natural killer cells and CD8+ T cells infiltration.In summary, the 14 features classifier accurately predicted robust SV gene pairs for cancer. The six SV pairs signature could predict resistance to ICI.© 2023. The Author(s).