diff --git a/_config.yml b/_config.yml index 89247507d3839..a9c59deeff18a 100644 --- a/_config.yml +++ b/_config.yml @@ -10,7 +10,7 @@ locale : "en-US" title : "Changwen Xu" title_separator : "-" name : &name "Changwen Xu" -description : &description "Graduate Student at Carnegie Mellon University" +description : &description "Ph.D. Student at University of Michigan" url : https://ChangwenXu98.github.io # the base hostname & protocol for your site e.g. "https://mmistakes.github.io" baseurl : "" # the subpath of your site, e.g. "/blog" repository : "academicpages/academicpages.github.io" @@ -82,12 +82,12 @@ analytics: author: name : "Changwen Xu" avatar : "profile.jpg" - bio : "Graduate Student at Carnegie Mellon University" - location : "Pittsburgh, PA" + bio : "Ph.D. Student at University of Michigan" + location : "Ann Arbor, MI" employer : # pubmed : "https://www.ncbi.nlm.nih.gov/pubmed/?term=john+snow" googlescholar : "https://scholar.google.com/citations?user=GyVx78kAAAAJ&hl=en" - email : "changwex@andrew.cmu.edu" + email : "changwex@umich.edu" researchgate : "https://www.researchgate.net/profile/Changwen-Xu-3" uri : bitbucket : diff --git a/_pages/about.md b/_pages/about.md index b846407d281b1..e06017a03e297 100644 --- a/_pages/about.md +++ b/_pages/about.md @@ -10,7 +10,7 @@ redirect_from: About Me ------ -I'm going to be a PhD student in MechE at University of Michigan advised by [Prof. Viswanathan](https://www.andrew.cmu.edu/user/venkatv/index.html). Earlier, I received M.S. in Computational Materials Science and Engineering at Carnegie Mellon University and did research in [Mechanical and AI Lab](https://sites.google.com/view/barati) advised by [Prof. Farimani](https://www.meche.engineering.cmu.edu/directory/bios/barati-farimani-amir.html). Besides, I received B.Eng in Materials Science and Engineering at South China University of Technology. +I am a Ph.D. student in Mechanical Engineering at University of Michigan advised by [Prof. Viswanathan](https://www.andrew.cmu.edu/user/venkatv/index.html) working on Molecular Machine Learning. Earlier, I received M.S. in Computational Materials Science and Engineering at Carnegie Mellon University and did research in [Mechanical and AI Lab](https://sites.google.com/view/barati) advised by [Prof. Barati Farimani](https://www.meche.engineering.cmu.edu/directory/bios/barati-farimani-amir.html). Besides, I received B.Eng in Materials Science and Engineering at South China University of Technology. My research interest lies in combining Artificial Intelligence with interdisciplinary science and engineering problems. My current research focuses on implementing and improving deep learning in molecular property prediction and conformation generation to understand complex chemical systems. I believe that deep learning models are able to learn representations from data so that we can understand scientific problems from a data science view, which is a significant transformation to traditional research strategies. diff --git a/_pages/cv.md b/_pages/cv.md index 507ab4fc6b80b..fc5de2054f45f 100644 --- a/_pages/cv.md +++ b/_pages/cv.md @@ -13,11 +13,17 @@ You can download my [CV](http://ChangwenXu98.github.io/files/CV.pdf) (last updat Education ====== +* Ph.D. in Mechanical Engineering, University of Michigan, 2027 (expected) * B.E. in Materials Science and Engineering, South China University of Technology, 2021 * M.S. in Computational Materials Science and Engineering, Carnegie Mellon University, 2022 Work experience ====== +* Summer 2023: Machine Learning Summer Internship + * Redesign Science + * Duties included: Improving Machine-Learned Collective Variables with Energy-based Path Construction + * Supervisor: Dr. Andreas Mardt + * Summer 2022 and Spring 2023: Research Assistant * Carnegie Mellon University * Duties included: Developing AI for molecule property prediction diff --git a/_publications/2023-03-03-Denoise.md b/_publications/2023-03-03-Denoise.md deleted file mode 100644 index 74ea849a66079..0000000000000 --- a/_publications/2023-03-03-Denoise.md +++ /dev/null @@ -1,17 +0,0 @@ ---- -title: "Denoise Pre-training on Non-equilibrium Molecules for Accurate and Transferable Neural Potentials" -collection: publications -permalink: /publication/2023-03-03-Denoise -date: 2023-03-03 -venue: 'arXiv preprint' -authors: Y Wang, C Xu, Z Li, and A Barati Farimani, arXiv preprint, 2023 - ---- - -Paper available here - ---- - -**Abstract** - -Machine learning methods, particularly recent advances in equivariant graph neural networks (GNNs), have been investigated as surrogate models to expensive ab initio quantum mechanics (QM) approaches for molecular potential predictions. However, building accurate and transferable potential models using GNNs remains challenging, as the quality and quantity of data are greatly limited by QM calculations, especially for large and complex molecular systems. In this work, we propose denoise pre-training on non-equilibrium molecular conformations to achieve more accurate and transferable GNN potential predictions. Specifically, GNNs are pre-trained by predicting the random noises added to atomic coordinates of sampled non-equilibrium conformations. Rigorous experiments on multiple benchmarks reveal that pre-training significantly improves the accuracy of neural potentials. Furthermore, we show that the proposed pre-training approach is model-agnostic, as it improves the performance of different invariant and equivariant GNNs. Notably, our models pre-trained on small molecules demonstrate remarkable transferability, improving performance when fine-tuned on diverse molecular systems, including different elements, charged molecules, biomolecules, and larger systems. These results highlight the potential for leveraging denoise pre-training approaches to build more generalizable neural potentials for complex molecular systems. diff --git a/_publications/2023-06-30-Denoise.md b/_publications/2023-06-30-Denoise.md new file mode 100644 index 0000000000000..536716aa2d79c --- /dev/null +++ b/_publications/2023-06-30-Denoise.md @@ -0,0 +1,17 @@ +--- +title: "Denoise Pretraining on Nonequilibrium Molecules for Accurate and Transferable Neural Potentials" +collection: publications +permalink: /publication/2023-06-30-Denoise +date: 2023-06-30 +venue: 'Journal of Chemical Theory and Computation' +authors: Y Wang, C Xu, Z Li, and A Barati Farimani, J. Chem. Theory Comput., 2023 + +--- + +Paper available here + +--- + +**Abstract** + +Recent advances in equivariant graph neural networks (GNNs) have made deep learning amenable to developing fast surrogate models to expensive ab initio quantum mechanics (QM) approaches for molecular potential predictions. However, building accurate and transferable potential models using GNNs remains challenging, as the data are greatly limited by the expensive computational costs and level of theory of QM methods, especially for large and complex molecular systems. In this work, we propose denoise pretraining on nonequilibrium molecular conformations to achieve more accurate and transferable GNN potential predictions. Specifically, atomic coordinates of sampled nonequilibrium conformations are perturbed by random noises, and GNNs are pretrained to denoise the perturbed molecular conformations which recovers the original coordinates. Rigorous experiments on multiple benchmarks reveal that pretraining significantly improves the accuracy of neural potentials. Furthermore, we show that the proposed pretraining approach is model-agnostic, as it improves the performance of different invariant and equivariant GNNs. Notably, our models pretrained on small molecules demonstrate remarkable transferability, improving performance when fine-tuned on diverse molecular systems, including different elements, charged molecules, biomolecules, and larger systems. These results highlight the potential for leveraging denoise pretraining approaches to build more generalizable neural potentials for complex molecular systems. \ No newline at end of file diff --git a/files/acs.jctc.3c00289.pdf b/files/acs.jctc.3c00289.pdf new file mode 100644 index 0000000000000..53f7afff6d162 Binary files /dev/null and b/files/acs.jctc.3c00289.pdf differ