arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2203.04698cs.LGq-bio.QM
收藏

分子生成的基于得分的生成模型

Score-Based Generative Models for Molecule Generation

Julia Computing, Inc(Julia计算公司);Deep Forest Sciences Inc.(深林科学公司);Carnegie Mellon University(卡内基梅隆大学)

Dwaraknath Gnaneshwar, Bharath Ramsundar, Dhairya Gandhi, Rachel Kurchin, Venkatasubramanian Viswanathan

摘要:

生成模型的最新进展使得探索从头分子生成的设计空间变得更加容易。然而,流行的生成模型,如GANs和规范化流,分别面临着由于对抗性训练和架构约束而导致的训练不稳定性等挑战。基于得分的生成模型通过使用分数函数近似对对数概率密度的梯度进行建模来回避这些挑战,而不是直接对密度函数进行建模,并使用退火Langevin动力学对其进行采样。我们相信,基于得分的生成模型可以开辟新的机会,在分子生成由于其架构的灵活性,如取代评分函数与SE(3)等变模型。在这项工作中,我们通过测试基于分数的分子生成模型的有效性奠定了基础。我们对来自ZINC数据集的150万个样本的自引用嵌入字符串(SELFIES)表示进行训练,并使用Moses基准测试框架来评估一套指标上生成的样本。

英文摘要:

Recent advances in generative models have made exploring design spaces easier for de novo molecule generation. However, popular generative models like GANs and normalizing flows face challenges such as training instabilities due to adversarial training and architectural constraints, respectively. Score-based generative models sidestep these challenges by modelling the gradient of the log probability density using a score function approximation, as opposed to modelling the density function directly, and sampling from it using annealed Langevin Dynamics. We believe that score-based generative models could open up new opportunities in molecule generation due to their architectural flexibility, such as replacing the score function with an SE(3) equivariant model. In this work, we lay the foundations by testing the efficacy of score-based models for molecule generation. We train a Transformer-based score function on Self-Referencing Embedded Strings (SELFIES) representations of 1.5 million samples from the ZINC dataset and use the Moses benchmarking framework to evaluate the generated samples on a suite of metrics.

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。

↑