Generative Coarse-Graining of Molecular Conformations
Authors: Wujie Wang, Minkai Xu, Chen Cai, Benjamin K Miller, Tess Smidt, Yusu Wang, Jian Tang, Rafael Gomez-Bombarelli
ICML 2022 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | 5. Experiments In this section, we introduce our experimental settings and discuss both the quantitative and qualitative results. We first introduce the benchmark datasets and comprehensive metrics in Section 5.1 and Section 5.2, and then visualize and analyze the results in Section 5.3. |
| Researcher Affiliation | Academia | 1Massachusetts Institute of Technology, USA 2Mila Qu ebec AI Institute, Canada 3Universite de Montr eal, Canada 4University of California San Diego, USA 5University of Amsterdam, Netherlands 6HEC Montr eal, Canada 7CIFAR AI Chair, Canada. |
| Pseudocode | No | The paper describes the method and its components in prose and equations, but does not include explicitly labeled pseudocode or algorithm blocks. |
| Open Source Code | Yes | Code Availability A reference implementation can be found at https://github.com/wwang2/Coarse Graining VAE |
| Open Datasets | Yes | We evaluate our models on 2 datasets that have been used as benchmarks for data-driven CG modeling ((Wang & G omez-Bombarelli, 2019; Wang et al., 2019; Husic et al., 2020)): MD trajectories of alanine dipeptide and chignolin. |
| Dataset Splits | Yes | For alanine dipeptide and chignolin, we randomly select 20,000 and 10,000 conformations respectively from the data trajectories and perform 5-fold cross-validation to compute mean and standard errors of evaluation metrics. |
| Hardware Specification | No | The paper does not provide specific hardware details such as GPU/CPU models, memory, or cloud instances used for running experiments. |
| Software Dependencies | No | The paper mentions software like Pytorch, Py Emma, RDKIT, and MDTraj, but does not provide specific version numbers for these software dependencies crucial for reproducibility. |
| Experiment Setup | Yes | Table 4: hyperparameters used for CGVAE |