arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2202.01338cs.LGcs.AIcs.CLq-bio.BM
收藏

Regression Transformer:面向分子语言建模的并行序列回归与生成方法

Regression Transformer: Concurrent sequence regression and generation for molecular language modeling

IBM Research Europe(IBM欧洲研究院);ETH Zurich(苏黎世联邦理工学院)

Jannis Born, Matteo Manica

AI总结:

针对分子生成模型缺乏连续属性归纳偏置、可控性不足的问题,提出Regression Transformer多任务模型,通过交替训练同时实现精准属性回归与条件分子生成,性能优于传统专用方法,可支撑材料设计等领域的属性驱动分子探索。

摘要:

尽管生成模型在自然科学领域已取得显著进展,但其可控性仍颇具挑战。分子或蛋白质生成模型存在一个根本性的缺失维度,即能够反映目标连续属性的归纳偏置。为此,我们提出Regression Transformer(RT,回归Transformer)——一种将回归抽象为条件序列建模问题的全新方法。这催生了多任务语言模型的新范式,可无缝衔接序列回归与条件序列生成任务。\n我们通过充分实验证明,尽管采用标称尺度的训练目标,RT在小分子、蛋白质和化学反应的属性预测任务中,性能均与传统回归模型相当甚至更优。关键在于,为同一模型输入连续属性作为前缀,即可得到极具竞争力的条件生成模型,在子结构约束、属性驱动的分子生成基准测试中,表现优于专用方法。我们的二分法依托一种全新的交替训练方案实现,该方案使模型能够按期望属性修饰种子序列,例如优化反应产率。\n综上,RT是首个在生物化学领域同时擅长预测和生成任务的多任务模型。它尤其适用于化学或蛋白质空间的属性驱动局部探索,或可为材料设计领域的基础模型发展铺平道路。\n复现本文所有实验的代码可在以下地址获取:https://github.com/IBM/regression-transformer

英文摘要:

Despite significant progress of generative models in the natural sciences, their controllability remains challenging. One fundamentally missing aspect of molecular or protein generative models is an inductive bias that can reflect continuous properties of interest. To that end, we propose the Regression Transformer (RT), a novel method that abstracts regression as a conditional sequence modeling problem. This introduces a new paradigm of multitask language models which seamlessly bridge sequence regression and conditional sequence generation. We thoroughly demonstrate that, despite using a nominal-scale training objective, the RT matches or surpasses the performance of conventional regression models in property prediction tasks of small molecules, proteins and chemical reactions. Critically, priming the same model with continuous properties yields a highly competitive conditional generative model that outperforms specialized approaches in a substructure-constrained, property-driven molecule generation benchmark. Our dichotomous approach is facilitated by a novel, alternating training scheme that enables the model to decorate seed sequences by desired properties, e.g., to optimize reaction yield. In sum, the RT is the first report of a multitask model that concurrently excels at predictive and generative tasks in biochemistry. This finds particular application in property-driven, local exploration of the chemical or protein space and could pave the road toward foundation models in material design. The code to reproduce all experiments of the paper is available at: https://github.com/IBM/regression-transformer

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。

↑