Premium
Effect of Structural Descriptors on the Design of Cyclin Dependent Kinase Inhibitors Using Similarity‐based Molecular Evolution
Author(s) -
Kawai Kentaro,
Karuo Yukiko,
Tarui Atsushi,
Sato Kazuyuki,
Omote Masaaki
Publication year - 2020
Publication title -
molecular informatics
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.481
H-Index - 68
eISSN - 1868-1751
pISSN - 1868-1743
DOI - 10.1002/minf.201900126
Subject(s) - in silico , similarity (geometry) , biological system , fingerprint (computing) , binary number , quantitative structure–activity relationship , computer science , computational biology , artificial intelligence , chemistry , mathematics , biology , machine learning , biochemistry , arithmetic , image (mathematics) , gene
In this study, we evaluated the effect of structural descriptors on the in silico design of bioactive compounds. The authors have proposed a molecular design technique for designing new bioactive compounds. In this approach, known fragments are combined to generate new structures, which are evolved to increase the similarity to a known active compound. We generated the structure of CDK2 inhibitors using four descriptors (three binary fingerprints and a numerical vector) to evaluate the effect of descriptors on the molecular design. Subsequently, the physicochemical properties of the generated compounds were compared and evaluated from a similarity viewpoint. As a result, it was clarified that better structures can be generated by using descriptors consisting of numerical vectors rather than binary fingerprints. Moreover, the compound generated using the numerical vector or a long‐bit fingerprint resulted in favorable docking scores. Although binary fingerprints such as MACCS are widely used in this field, this result shows that it is important to use numeric vectors, or at least to use long‐bit fingerprints, to design drug‐like CDK2 inhibitors by the similarity‐based structure generation.