A Plug-and-Play Quaternion Message-Passing Module for Molecular Conformation Representation
Authors: Angxiao Yue, Dixin Luo, Hongteng Xu
AAAI 2024 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on various molecular datasets show that plugging our QMP module into existing invariant GNNs leads to consistent and significant improvements in molecular conformation representation and downstream tasks. |
| Researcher Affiliation | Academia | Angxiao Yue1, Dixin Luo2, Hongteng Xu1,3* 1Gaoling School of Artifical Intelligence, Renmin University of China, Beijing, China 2School of Computer Science and Technology, Beijing Institute of Technology, Beijing, China 3 Beijing Key Laboratory of Big Data Management and Analysis Methods, Beijing, China angxiaoyue@ruc.edu.cn, dixin.luo@bit.edu.cn, hongtengxu@ruc.edu.cn |
| Pseudocode | No | The paper describes the QMP module and its steps in prose and with mathematical formulas, but it does not include a clearly labeled 'Pseudocode' or 'Algorithm' block. |
| Open Source Code | Yes | The code is available at https://github.com/Angxiao Yue/QMP. |
| Open Datasets | Yes | We conduct experiments on three datasets, including MD17 (Chmiela et al. 2017), MD17@CCSD (Chmiela et al. 2018), and OC20 (Chanussot et al. 2021b). |
| Dataset Splits | Yes | We train each model using 1,000 conformations and 1,000 for validation, with the remaining used for testing. |
| Hardware Specification | Yes | To ensure experimental fairness, we conduct the experiments using the same hyperparameter settings and hardware environment (NVIDIA Ge Force RTX 3090). |
| Software Dependencies | No | The paper mentions using 'DIG2 framework (Liu et al. 2021a)' and 'Open Catalyst Project (OCP) framework3 (Chanussot et al. 2021a)' but does not specify version numbers for these frameworks or any other software dependencies like Python, PyTorch, etc. |
| Experiment Setup | Yes | To ensure experimental fairness, we conduct the experiments using the same hyperparameter settings and hardware environment (NVIDIA Ge Force RTX 3090). |