基于预筛选的海量分子结构检索算法

Retrieval algorithm of large amounts of molecular structure based on pre-screening

针对大数据环境下,化学分子结构检索低效、通用图同构VF2算法检索量过大的缺陷,提出了基于预筛选的ASVF2算法,建立了基于分布式的分子检索模型。实验结果表明,在包含20万个化学结构的数据中,该方法可以快速准确地检索包含特定信息的分子,大幅降低了分子结构检索的复杂度,模型具有稳定的可扩展性。

In view of the problems that chemical molecular structure retrieved is inefficient in large data environment, and the defect of the vast retrieval of the general graph isomorphism VF2 algorithm, an algorithm named ASVF2 algorithm was proposed based on prescreening. At the same time, a distributed molecular retrieval model was established. The experimental results show that in the chemical data which include 200 thousand structures of compound, this method can rapidly and accurately retrieve molecules comprising specific information, and greatly reduce the complexity of the molecular structure of retrieval, Besides, the model is stable scalability.