Polymer graph neural networks for multitask property learning

作     者：Owen Queen Gavin A.McCarver Saitheeraj Thatigotla Brendan P.Abolins Cameron L.Brown Vasileios Maroulas Konstantinos D.Vogiatzis 

作者机构：Department of MathematicsUniversity of TennesseeKnoxvilleTN 37996-1320USA Department of Electrical Engineering and Computer ScienceUniversity of TennesseeKnoxvilleTN 37996-2250USA Department of ChemistryUniversity of TennesseeKnoxvilleTN 37996-1600USA Eastman Chemical CompanyKingsportTN 37660USA 

出 版 物：《npj Computational Materials》 (计算材料学（英文）)

年 卷 期：2023年第9卷第1期

页      面：1420-1429页

核心收录：

学科分类：081704[工学-应用化学] 07[理学] 081203[工学-计算机应用技术] 08[工学] 0817[工学-化学工程与技术] 070305[理学-高分子化学与物理] 080501[工学-材料物理与化学] 0805[工学-材料科学与工程（可授工学、理学学位）] 0835[工学-软件工程] 0703[理学-化学] 0812[工学-计算机科学与技术（可授工学、理学学位）] 

基　　金：This research is generously supported by Eastman Chemical Company grant no.EMN-20-F-S-01.We also acknowledge the Infrastructure for Scientific Applications and Advanced Computing(ISAAC)of the University of Tennessee 

主　　题：properties space polyester 

摘      要：The prediction of a variety of polymer properties from their monomer composition has been a challenge for material informatics,and their development can lead to a more effective exploration of the material *** this work,POLYMERGNN,a multitask machine learning architecture that relies on polymeric features and graph neural networks has been developed towards this *** provides accurate estimates for polymer properties based on a database of complex and heterogeneous polyesters(linear/branched,homopolymers/copolymers)with experimentally refined *** POLYMERGNN,each polyester is represented as a set of monomer units,which are introduced as molecular graphs.A virtual screening of a large,computationally generated database with materials of variable composition was performed,a task that demonstrates the applicability of the POLYMERGNN on future studies that target the exploration of the polymer ***,a discussion on the explainability of the models is provided.