Structure-aware conditional variational auto-encoder for constrained molecule optimization

作者：

Highlights：

• Small modifications in molecule structures potentially lead to drastic increasing of properties.

• Graph alignment can discover the similarity of molecule pairs.

• Two-level generation improves the performance of constrained molecule optimization.

• Variational evidence lower bound facilitates the constrained molecule optimization with structure similarity.

摘要

•Small modifications in molecule structures potentially lead to drastic increasing of properties.•Graph alignment can discover the similarity of molecule pairs.•Two-level generation improves the performance of constrained molecule optimization.•Variational evidence lower bound facilitates the constrained molecule optimization with structure similarity.

论文关键词：Molecule optimization,Conditional generation,Drug discovery

论文评审过程：Received 10 January 2021, Revised 28 December 2021, Accepted 7 February 2022, Available online 9 February 2022, Version of Record 12 February 2022.

论文官网地址：https://doi.org/10.1016/j.patcog.2022.108581