Polymer informatics is being utilized to accelerate polymer discovery. However, the practical realization of the designed polymer is still slow due to synthesis challenges, e.g., difficulties with the identification of potential polymerization mechanisms and optimal reactants/solvents/processing conditions. In the past, synthesis pathways adopted for a target polymer have been heavily dependent on chemical intuition and past experience. To expedite this process, we have developed a data-driven approach to assist in polymer retrosynthesis planning. In this work, a dataset of polymerization reactions was manually accumulated from various resources to extract hundreds of synthetic templates and used as the training set. Further, a similarity metric was adopted to select synthetic templates and similar existing reactants for the new target polymer. Finally, prediction accuracy was measured by comparison with ground truth and/or bench chemists' estimation. The proposed data-driven polymer synthesis recommendation model has been deployed at https://www.polymergenome.org.

中文翻译：

---

数据辅助的聚合物逆合成规划

聚合物信息学正被用来加速聚合物的发现。然而，由于合成挑战，例如难以确定潜在聚合机制和最佳反应物/溶剂/加工条件，设计聚合物的实际实现仍然很慢。过去，目标聚合物采用的合成途径在很大程度上依赖于化学直觉和过去的经验。为了加快这一过程，我们开发了一种数据驱动的方法来协助聚合物逆合成规划。在这项工作中，从各种资源中手动积累了聚合反应的数据集，以提取数百个合成模板并用作训练集。此外，采用相似性度量为新目标聚合物选择合成模板和相似的现有反应物。最后，通过与真实情况和/或实验室化学家的估计进行比较来衡量预测准确性。提议的数据驱动的聚合物合成推荐模型已部署在 https://www.polymergenome.org。