Лаборатория № 22 ИВС РАН «МНОГОМАСШТАБНОЕ ЭКСПЕРИМЕНТАЛЬНОЕ ИССЛЕДОВАНИЕ И МОДЕЛИРОВАНИЕ ПОЛИМЕРНЫХ КОМПОЗИТОВ НА ОСНОВЕ ПЕРСПЕКТИВНЫХ ТЕРМОПЛАСТОВ ДЛЯ ПРОМЫШЛЕННОГО ПРИМЕНЕНИЯ» (Лаборатория «ПОЛИКОМП», PolyCompLab).
To browse data in PolyAskInG the PolyAskInG_View program could be used. PolyAskInG_View is compatible only with 64-bit Windows. The program is distributed in 7z archive and could be launched right after extraction. No installation required.
To open database in PolyAskInG_View use “File -> Open Database” menu. In the window opened choose the database file extracted from the archive downloaded from this page. On database opening it will be possible to browse structures and properties of polyimides.
Also, one may filter structures by glass transition temperature values or monomer unit molecular weight. To do this use “Filter” menu options. In the window opened enter the minimum and maximum values of Tg or Mw and press “OK” button to filter data. To reset filter, just open window once again and press “OK” button without changing values in the input fields.
Using “Export” menu item it is possible to save structure of polyimide shown in the program windows in MOL file format.
Both PolyAskIng database and PolyAskInG_View program are distributed compressed by 7z software that could be obtained from https://www.7-zip.org/ .
To facilitate an active development in this area, we make our PolyAskInG database, as well as the source code of the developed GCNN models, free of use for non-profit purposes in the case of basic research activity (reference to corresponding publication is obligatory). The source code is available on https://github.com/polycomplab.
For the reference purpose and for the details of PolyAskIng database generation, please, see
I.V. Volgin, P. Batyr, A.V. Matseevich, A.Y. Dobrovskiy, M.V. Andreeva, V.M. Nazarychev, S.V. Larin, M.Ya. Goikhman, Yu.V. Vizilter, A.A. Askadskii, S.V. Lyulin. Machine Learning with Enormous “Synthetic” Datasets: Predicting Glass Transition Temperature of Heterocyclic Polymers Using Graph Convolutional Neural Networks. 2022.