抗体药物偶联物 (ADC) 已成为一类有前景的靶向癌症治疗药物。进一步的改进对于释放其全部潜力至关重要，目前其潜力因缺乏经过验证的目标和有效载荷而受到限制。开发有效 ADC 的基本方面涉及表面抗原的识别，理想地将目标肿瘤细胞与健康类型区分开来，均匀表达，并伴有能够选择性靶向的高效有效负载。在这项研究中，我们利用 Lantern Pharma 专有的 AI 平台药物定位和救援响应算法 (RADR®) 集成了来自 Human Protein Atlas、Xenabrowser 和 Gene Expression Omnibus 的转录组学、蛋白质组学、免疫组织化学和细胞表面膜数据集。我们将其与基于证据的过滤相结合来识别具有改进的肿瘤选择性的 ADC 靶点。我们的分析确定了一组 82 个靶标和总共 290 个靶标适应症组合，用于有效的肿瘤靶向。我们通过针对 22 种肿瘤亚型查询 TCGA 突变数据库中的 416 个基因，评估了肿瘤突变对靶标表达水平的影响。此外，我们还使用 NCI 开发治疗计划编制了一份化合物目录，以识别潜在的有效负载。我们的有效负载挖掘策略根据从 pM 到 10 nM 范围的 GI50 值，结合 9 种不同癌症适应症的敏感性模式，将 729 种化合物分为三个子类。我们的结果确定了各种不同的目标和有效载荷，这可以为精确 ADC 定位提供多种选择。我们提出了一种初步方法来确定合适的目标-指示-有效负载组合，作为开发未来 ADC 候选者的宝贵起点。版权所有：© 2024 Kathad 等人。这是一篇根据知识共享署名许可条款分发的开放获取文章，允许在任何媒体上不受限制地使用、分发和复制，前提是注明原始作者和来源。

Antibody-Drug Conjugates (ADCs) have emerged as a promising class of targeted cancer therapeutics. Further refinements are essential to unlock their full potential, which is currently limited by a lack of validated targets and payloads. Essential aspects of developing effective ADCs involve the identification of surface antigens, ideally distinguishing target tumor cells from healthy types, uniformly expressed, accompanied by a high potency payload capable of selective targeting. In this study, we integrated transcriptomics, proteomics, immunohistochemistry and cell surface membrane datasets from Human Protein Atlas, Xenabrowser and Gene Expression Omnibus utilizing Lantern Pharma's proprietary AI platform Response Algorithm for Drug positioning and Rescue (RADR®). We used this in combination with evidence based filtering to identify ADC targets with improved tumor selectivity. Our analysis identified a set of 82 targets and a total of 290 target indication combinations for effective tumor targeting. We evaluated the impact of tumor mutations on target expression levels by querying 416 genes in the TCGA mutation database against 22 tumor subtypes. Additionally, we assembled a catalog of compounds to identify potential payloads using the NCI-Developmental Therapeutics Program. Our payload mining strategy classified 729 compounds into three subclasses based on GI50 values spanning from pM to 10 nM range, in combination with sensitivity patterns across 9 different cancer indications. Our results identified a diverse range of both targets and payloads, that can serve to facilitate multiple choices for precise ADC targeting. We propose an initial approach to identify suitable target-indication-payload combinations, serving as a valuable starting point for development of future ADC candidates.Copyright: © 2024 Kathad et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.