该课题组设计了一个主动的深度学习框架，利用组学来实现可扩展的、可优化的识别诱导复杂表型的化合物。他们的可推广算法在经典召回方面优于最先进的模型，在两次血液学发现活动中转化为13-17倍的表型命中率增加。将该算法与循环实验室签名细化步骤相结合，小组将命中率和分子洞察力提高了两倍。总之，他们的框架使有效的表型命中识别运动，具有广泛的潜力，以加速药物发现。
据悉，表型药物筛选仍然受到巨大的化学空间和扩展实验工作流程的技术挑战的限制。为了克服这些障碍，已经开发了计算方法来确定化合物的优先级，但它们要么依赖于缺乏泛化性的单任务模型，要么依赖于抵制优化的启发式基因组代理。
附：英文原文
Title: Active learning framework leveraging transcriptomics identifies modulators of disease phenotypes
Author: Benjamin DeMeo, Charlotte Nesbitt, Samuel A. Miller, Daniel B. Burkhardt, Inna Lipchina, Doris Fu, Peter Holderreith, David Kim, Sergey Kolchenko, Artur Szalata, Ishan Gupta, Christine Kerr, Thomas Pfefer, Raziel Rojas-Rodriguez, Sunil Kuppassani, Laurens Kruidenier, Parul B. Doshi, Mahdi Zamanighomi, James J. Collins, Alex K. Shalek, Fabian J. Theis, Mauricio Cortes
Issue&Volume: 2025-10-23
Abstract: Phenotypic drug screening remains constrained by the vastness of chemical space and technical challenges scaling experimental workflows. To overcome these barriers, computational methods have been developed to prioritize compounds, but they rely on either single-task models lacking generalizability or heuristic-based genomic proxies that resist optimization. We designed an active deep-learning framework that leverages omics to enable scalable, optimizable identification of compounds that induce complex phenotypes. Our generalizable algorithm outperformed state-of-the-art models on classical recall, translating to a 13-17x increase in phenotypic hit-rate across two hematological discovery campaigns. Combining this algorithm with a lab-in-the-loop signature refinement step, we achieved an additional two-fold increase in hit-rate and molecular insights. In sum, our framework enables efficient phenotypic hit identification campaigns, with broad potential to accelerate drug discovery.
DOI: adi8577
Source: https://www.science.org/doi/10.1126/science.adi8577