SEARCH
検索詳細GANTZER PHILIPPE ANDRE JOSEPH (ガンツアー フイリペ アンドレ ジヨセフ)
| 総合イノベーション創発機構化学反応創成研究拠点 | 特任助教 |
研究者基本情報
■ 学位■ URL
researchmap URLホームページURL■ ID 各種
研究者番号
- 80972611
研究キーワード研究分野
研究活動情報
■ 論文- Integration of DOPtools and CADS in a Web-Based User Interface for Structural Descriptor Calculation, Model Optimization, and Prediction
Philippe Gantzer; Micke Kuwahara; Keisuke Takahashi; Pavel Sidorov
Journal of Chemical Information and Modeling, 2026年03月19日, [査読有り], [筆頭著者, 責任著者]
英語, 研究論文(学術雑誌) - SynPlanner: An End-to-End Tool for Synthesis Planning
Tagir Akhmetshin; Dmitry Zankov; Philippe Gantzer; Dmitry Babadeev; Anna Pinigina; Timur Madzhidov; Alexandre Varnek
Journal of Chemical Information and Modeling, 2025年01月13日, [査読有り]
英語, 研究論文(学術雑誌) - DOPtools: a Python platform for descriptor calculation and model optimization
Said Byadi; Philippe Gantzer; Timur Gimadiev; Pavel Sidorov
Digital Discovery, 2025年, [査読有り]
研究論文(学術雑誌) - Chemography‐guided analysis of a reaction path network for ethylene hydrogenation with a model Wilkinson's catalyst
Philippe Gantzer; Ruben Staub; Yu Harabuchi; Satoshi Maeda; Alexandre Varnek
Molecular Informatics, 2025年01月, [査読有り], [筆頭著者]
研究論文(学術雑誌) - Challenges for Kinetics Predictions via Neural Network Potentials: A Wilkinson’s Catalyst Case
Ruben Staub; Philippe Gantzer; Yu Harabuchi; Satoshi Maeda; Alexandre Varnek
Molecules, 2023年05月31日, [査読有り]
研究論文(学術雑誌) - Development and comparison of inverse-QSPR approaches
Philippe Gantzer
2021年11月04日, [筆頭著者]
フランス語, 学位論文(博士), The use of Quantitative Structure-Property Relationships (QSPR) has become frequent to quickly predict molecules’ properties. Also, new chemical structures can be designed nowadays by virtual generation. The inversion of QSPR models (i-QSPR) is one of the existing virtual generation methods, which allows to obtain molecules with desired properties. During this PhD work, we first focussed on implementing and improving such methods. Following a study of the literature, three methods, based on fragments assemblies (F), successive modifications of genetic graphs (G) and a variational autoencoder (E), were selected to supplement our database. Our database gathers 785 hydrocarbons and oxygenated molecules, with their associated flash point (FP) value. Up to 5 million structures were generated by each method. Then, a method for comparing generation methods was proposed. Our approach analyses the distribution of molecules generated by each algorithm, in a small-dimensional space. The F and G methods were compared with this new tool, and we highlighted the best performances of G to complete our FP database in a diverse and representative way. For the generation of structures responding to a constraint on the FP value, we set up a selection of the number of molecules to be modified by G. The restriction of the number of molecules to be modified by G proves to be effective for the generation of molecules with a FP value greater than 400 K; while it degrades the performance of generation of molecules with a FP value less than 400 K. - Comparisons of Molecular Structure Generation Methods Based on Fragment Assemblies and Genetic Graphs
Philippe Gantzer; Benoit Creton; Carlos Nieto-Draghi
Journal of Chemical Information and Modeling, 61, 9, 4245, 4258, American Chemical Society ({ACS}), 2021年09月27日, [査読有り], [筆頭著者]
英語, 研究論文(学術雑誌) - Modelling of ready biodegradability based on combined public and industrial data sources
F. Lunghini; G. Marcou; P. Gantzer; P. Azam; D. Horvath; E. Van Miert; A. Varnek
SAR and QSAR in Environmental Research, 31, 3, 171, 186, Informa {UK} Limited, 2020年03月03日, [査読有り]
研究論文(学術雑誌) - Inverse‐QSPR for de novo Design: A Review
Philippe Gantzer; Benoit Creton; Carlos Nieto‐Draghi
Molecular Informatics, Wiley, 2019年11月, [査読有り], [筆頭著者]
研究論文(学術雑誌)
