arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2401.11403cs.LGcs.CLq-bio.BM
收藏

MolTailor:通过文本提示定制化学分子表示以适配特定任务

MolTailor: Tailoring Chemical Molecular Representation to Specific Tasks via Text Prompts

Harbin Institute of Technology(哈尔滨工业大学)

Haoqiang Guo, Sendong Zhao, Haochun Wang, Yanrui Du, Bing Qin

AI总结:

MolTailor利用语言模型作为智能体,根据任务文本描述定制分子表示,突出任务相关特征,提升预测性能,优于基线方法。

摘要:

深度学习现已广泛应用于药物发现领域,显著加速了研发进程并降低了成本。作为最基础的构建模块,分子表示对于预测分子性质以支持各种下游应用至关重要。现有大多数方法试图整合更多信息以学习更好的表示。然而,并非所有特征对于特定任务都同等重要。忽视这一点可能会损害训练效率和预测准确性。为解决这一问题,我们提出了一种新颖方法,将语言模型视为智能体,将分子预训练模型视为知识库。该智能体通过理解任务的自然语言描述,在分子表示中突出与任务相关的特征,正如裁缝为客户定制衣物一般。因此,我们将此方法命名为MolTailor。评估结果表明,MolTailor在性能上优于基线方法,验证了增强分子表示学习相关性的有效性。这展示了语言模型引导优化的潜力,能够更好地利用和释放现有强大分子表示方法的能力。我们的代码可在https://github.com/SCIR-HI/MolTailor获取。

英文摘要:

Deep learning is now widely used in drug discovery, providing significant acceleration and cost reduction. As the most fundamental building block, molecular representation is essential for predicting molecular properties to enable various downstream applications. Most existing methods attempt to incorporate more information to learn better representations. However, not all features are equally important for a specific task. Ignoring this would potentially compromise the training efficiency and predictive accuracy. To address this issue, we propose a novel approach, which treats language models as an agent and molecular pretraining models as a knowledge base. The agent accentuates task-relevant features in the molecular representation by understanding the natural language description of the task, just as a tailor customizes clothes for clients. Thus, we call this approach MolTailor. Evaluations demonstrate MolTailor's superior performance over baselines, validating the efficacy of enhancing relevance for molecular representation learning. This illustrates the potential of language model guided optimization to better exploit and unleash the capabilities of existing powerful molecular representation methods. Our code is available at https://github.com/SCIR-HI/MolTailor.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑